Cypress Semiconductor Corporation (NASDAQ:CY)
November 09, 2011 12:00 pm ET
Unknown Speaker -
Thomas H. Werner - Chairman, Chief Executive Officer and President
Simon Segars - Executive Vice President, General Manager of Physical IP Division and Director
Christopher A. Seams - Executive Vice President of Sales and Marketing
Dana Nazarian - Executive Vice President of Memory and Imaging Division
Badri Kothandaraman -
T. J. Rodgers - Co-Founder, Chief Executive Officer, President, Director, Director of Cypress Envirosystems, Director of Agiga Tech, Director of Bloom Energy and Member of Board of Trustees at Dartmouth College
Brad W. Buss - Chief Financial Officer, Principal Accounting Officer, Executive Vice President of Finance & Administration and Corporate Secretary
Norman P. Taffe - Executive Vice President of Consumer & Computation Division
Jeff Eastman -
Dinesh Ramanathan - Executive Vice President of Data Communications Division
Unknown Executive -
Tim Olson -
Leon Farasati -
Unknown Analyst -
Brad W. Buss
So again, thanks for coming to our annual meeting. Pretty excited for you all to be kind of be here. You got the doughnuts. Great, thank you. So again for those of you on the webcast, we have a lot of people on the webcast. We just want you to know, there's good body language everywhere, especially mine, because there's a lot of body to have language with. And I wanted to make sure you all have that visual experience.
The standard Safe Harbor, right. So it is Cypress, it is the analyst day and there will be a lot of interesting things that will be coming out, generally from my supervisor, right on queue, thank you very much. We encourage you to take a look through the 10-Q, which was recently filed. Always a lot of good tidbits in there. Here's kind of the overall agenda that we're going to be going through. I'll do a kind of a quick overview on some financial base setting. T.J. will comment and kind of give his thoughts on the direction of where we're heading and then do an intro of our newest member, Tim Olson, who is officially non-stealth now and we also have Tom Werner, who is part of this venture as well, he'll be kicking that off. And then we'll go through our major product line. So we used to kind of have a reference internally on the 3 amigos, and we actually have 4 amigos now running some of the product groups. You're going to see all 4 of them up here and I'll take you through a little bit of that so you can understand the organization that were there. But it'll be the big things, SRAM, USB, our PSoC family, and then also our touchscreen and user interface group. And then we'll wrap it up with a very detailed Q&A. The demos will all be out here. We'll have a light lunch if some of you want it or if you got to blast off to get ready for calls, you can do that as well. And everything is being webcast, obviously, so that's important for everyone to know.
So just kind of quickly, we want to keep all questions at the end. If you could silence your phones, that would be great. Any weird ring tone tap and its $20 donation to Second Harvest Food Bank, which is kind of the host charity here.
Thailand, we got a lot of questions on that. We have no direct exposure in there, we don't really have any major subcons or anything in there. Obviously, we're concerned on the poor people there and customers and then impacts on their supply chain. There's nothing that we're seeing at this point but we're keeping our eye pretty close on that. And then the new division structure that I mentioned, here's kind of how we're splitting up internally. So you're going to see the 4 Amigos, as I call them here on the left, right here. Dana's running the memory group, so basically the Async and the SRAM as you all know. DCD under Badri, who is the newest of the Amigos, and welcome to the club, Badri, it's been exciting time. This is of your duties you have to do. And that's really focus 100% on USB, and obviously the West Bridge and all the derivatives are part of that, and he'll give you a detailed talk on what's going on there, some pretty exciting stuff. So Norm, who is kind of the old CCD from you guys, is focusing 100% now on TrueTouch and our Ovation, our optical FingerNav and our trackpad solutions.
And then PSD, our programmable solutions division is a new name for you guys, and Dinesh is going to be taking that over. Dinesh used to be here. We kind of popped him over here, and he's managing all of our PSoC family. So 1, 3, 5, all the features that will be coming out of that. And then we have our Emerging Tech division, there's really no big change in there, AgigA Tech, Delta Tech, and Cypress Enviro, et cetera, okay? And we're looking at changing the external reporting that you see. We don't have that 100% defined. We'll be kicking that off really probably in Q1 of next year and then we'll obviously restate all the information and put a bunch of data on the website, so you can manage your model.
So just kind of on the growth proposition, and again, most of you have seen at proprietary solutions, that is really what we're all about. That's what you're seeing in the demos, that's what really you're going to be hearing in the vast majority of the presentations. Big focus on mobile phones, right? I mean, it's our biggest end market right now. It's close to 30% of what we're doing, and highly concentrated in the smartphone area. And obviously a big leader in touchscreen. You're going to hear touch, touch, touch, touch everywhere. And it's touching UI, right? I mean, we're going at that. We are really the UI guy of choice. 100% programmable and no one else is doing that and we keep pushing that lead further. PSoC, I think most of you know its huge tam, we're #1 in SRAM, we're #1 in the USB, and those will keep getting better between consolidation and new products that are getting rolled out. And we're starting to see the Emerging Tech guys start to ramp. We also have Tim, who will be giving us both 2 revenue down the road. Right, Tim? Who else would I like to see? Get him on line. And the balance sheet's good, and as you know we've been returning pretty well. So I'm not going to dwell on of the historical stuff, but we came pretty nicely off the bottom. We've been managing revenues that are much higher than the prior downturn. Our GP, is we used to talk about a 50% model, we're up around a 60% range now. And again, remember the prior downturn, we were sub50 into the high 39s when things really hit. We're nowhere near those levels now because of the FlexFab that we're doing.
OpEx, we've done a pretty good job on it. We've got very good kudos from most of the one that [ph], and that is going to continue. We are reallocating a lot of our expense into the faster growing areas. We've moved stuff offshore. We've become more efficient. We're very focused on OpEx and becoming leaner and more efficient in how we do everything, whether it's R&D, sales and marketing, So you’re going to see that A, nothing is sacred and a continued big focus IN the cost of goods sold area. You're going to see that continue to grow and that's going to drive a lot of leverage going forward. We use to kind of talk in the 20% range, right? And even before the prior downturn, we were barely hitting our head on it. Now we're up kind of in the mid-20s, and our model is still up around 30, and I'll take you through that.
EPS. Again, it's been very strong and even with the pullback in revenue in Q4, right? We're at levels that were far better than we've seen over the last couple of years and I think that strength will continue. Very low cost -- OpEx, before I came here, we were regularly spending $100 million, $150 million a year, and now with our FlexFab, we're in these $30 million to $50 million ranges and I kind of split '11 out because we did the estimate expansion to support all the PSoC growth, which was definitely needed and then our little friend Tim here is spending a few bucks and he'll tell you about that, so I really look at this as kind of a one-off in this year and I'd expect the core business to continue to trend around this range and then we may have a little blip here and there on the odds investment that we make, so driving very good cash flow. Most of you know we finished the $600 million buyback that worked out pretty well. We launched at $400 million buyback. Baby buyback number 2 as Neil and I call it, and this kind of gives you a perspective on the number of shares that could take a note out of the variety of prices and basically the impact that it would do to the share count. So we're very committed to returning a lot of the excess cash back to you. Buybacks, when they're opportunistic and we think they make sense and obviously, the dividend that we've kicked off is the paying a pretty nice yield to you guys.
So again this is just the guidance recap that I did before. Really no change in there, other than on the TrueTouch end of it. We said that we expect it to trend to the higher end of the 230 to 250. We expect it to be north of 250 and Norm will talk a little further on TrueTouch and where we're going with that, so we're pretty pleased to be able to turn in that number. It's a hell of an increase year-on-year, and we've got a very nice market share in there. As far as what's going on kind of quarter-to-date, I think all of the bookings backlog, all the information we're getting from our customers and distributors support the guidance that we basically have given out, so we're very comfortable with that. And I think they're still very choppy. I'm sure we'll talk about it later. The distributors continue to remain cautious, things are choppy in certain end markets but I don't think the world is falling off and I'm not sure we're at the bottom yet or not, but I think it's not too far off.
The model again, again, just to reiterate, we're looking to grow 2x of the industry. We think our core GM should run around the 58 to 60 to allow us to grow and also play in some of the markets that we're playing in. OpEx around 30, the TBT around 30. And again, we're continuing to invest very heavily. I mean, we are rolling out platform products like there's no tomorrow. We're already working on the next gen of a lot of those and we're investing in these Emerging Tech, so we are not laying off the investment. Trust me. There's a ton of meetings on spending R&D money. Right, Paul? And again, like I said, we're going to continue to drive the cash back and we continue to hope to outperform the market, so, well, this was prior to today, by the way, so it kind's of a daily thing. We've been doing a lot better than the average, a lot better than our peer semiconductor benchmarks, but I called it decent because I don't get too excited on 5%. The 3%, we're very proud of, that's really all post of SunPower spend days as well, so we've been outperforming the market. The 5 years have been spectacular, I mean, it's been pretty tough, and now we're paying dividend on top of this, a little extra kicker for everybody.
So just a quick commercial on the Emerging Tech, right? We've been seeing that continue to grow. We have different companies in there at different levels of either production or design and/or revenue. We started to see some nice contributions this year and I expect to see that continue up next year into kind of the $50-plus million range. So again, starting to become a little more meaningful. And very different model I think, as most of you know, I don't think there's another company out there that really believes in this model and/or has it honed like the way we do. And it all started with T.J., dozens of years ago. And we've had a pretty good history. We've had a couple little things that quite haven't been too exciting, right, T.J.? But overall, most of those have done well. We have friends at PSoC still here, even though they've left our campus for bigger, nicer areas. We still love them. And PSoC's been a huge success and we're expecting to see Deca Tech and AgigA and Enviro and a couple of the others do as well, right? So we're pretty pleased with where that's gone. I think hopefully north of $50 million, below choppy in some of the macro. They're all getting good design win opportunities, which is encouraging. And I'm hoping it breaks even by Q3 '12. I mean, it's been confident at $0.45 a quarter in EPS, so it's a pretty decent nut and we'll start to get some benefits from that. And I think the big question for all of you to look at is, I think a lot of you kind of look at the consolidated. You really need to look at the core business and understand that and then value the emerging tax really kind of separately. And I think there could be a $3 to $5 impact on that down the road.
Okay. And it wouldn't be Cypress above something else, so we kind of have stealth, and the SunPower, so now I'm looking at stealth 2.0, it's kind of the new one. We're kind of looking at the disruptive investment. We're not going to talk on what it is. Again its semiconductor related. It's leveraging a lot of what we do and what we're good at. And we're still working on a deal structure it may not come into fruition just kind of doing my forward-looking language stuff. And if it does, it's something I'd expect to be done in the oven by Q1 of 12, and if it makes sense at that time we'll share that with you. All right? So with that, I'm going to turn it over to our CEO, T.J. Rodgers.
T. J. Rodgers
Okay, first of all, since all meetings like this have potential legal ramifications, I want to warn you all not to believe a word that my Chief Financial Officer said to you. This is what he looks like when you're not here. Sure, yes, it wasn't Friday afternoon. Every Friday afternoon it was a special one-time event. Like whenever we lose money, special one-time-event when we make money the machine actually, right, okay. You've been warned, right? Okay.
Well, here's what we're here to talk about. This is a picture taken next to my office this morning. You can see my Green Bay coat, my Green Bay tie. I'm getting all warm, so I'm taking it off, showing my Green Bay shirt, my Green Bay socks and believe it or not I actually do own, but I'm not it wearing today, Green Bay underwear boxer type, which I don't like, I mean, it's a long one if you have boxers. This is last year, Packers beating Steelers in the Super Bowl, surprise. This is this year, beating the Saints in the first game, that's a rookie. I actually have that picture near my desk so I can point to the managers who come in at 5:00 in Friday night making the deadline for whatever is due to get into my desk, so they realize what they look like to me. Then they beat the Panthers. Here's my boy, Clay Matthews, the guy that was in the picture. Then they beat the Bears, and of course, at 187 games, this is the oldest rivalry in the NFL. There's Clay again taken down is Rick's forte. This is my favorite of all the pictures. This is the Bear quarterback and you can see the Wolves hanging on his leg and this is a linebacker taking my perfectly legal [indiscernible] helmet, shoulder pad to shoulder pad. You also noticed what I love -- notice the fingers are kind of going limp here, there's like no life left in them. And this guy is not going to fly backwards, he's simply going to drop and hid directly in the bottom of this picture. And they beat the Falcons in Atlanta. A team I've come not to like. I think they're not good sports. This is their old uniform, brown helmets to signify the leather. These are their 1929 30-31 uniforms. This is a -- in memory of their 1931 championship. Team Green Bay's won 13 NFL championships, more than any other teams since the league was formed. In those days, they were blue and gold, not green and gold, and they were blue and gold because Curly Lambeau [ph], their first coach and founder played at Notre Dame. He played for George Gibb [ph], Or he played for Newt Rockney [ph] with George Gibb [ph] at Notre Dame. They beat the Rams. Here they're beating the Vikings. That's always one of my favorites because the Vikings are the heated rivals. In that game Charles Woodson had 2 interceptions and in 1 scene it was absolutely great, it was on television. He took the interception found a Packer fan and gave him the ball. Not bad. And of course, they just beat San Diego. I don't have pictures up yet. And they're on the cover of Sports Illustrated, the story being, common boy makes it Green Bay with a killer quarter receiver, and it's really true. By the way, Aaron Rodgers, R-O-D-G-E-R-S, just like me. He is from Chico, California. His mother is a hairdresser and her name is Darla. He actually belongs in the Green Bay.
Okay. A little bit on stock before I do the main event which is to introduce Deca Tech. At 2011 year-to-date, on this graph I've compared Cypress, the NASDAQ in purple, the Dow Jones in green, which has fared better than the NASDAQ this year. I've also put in stocks in black, so if you wanted to compare us, you compare blue to purple, NASDAQ, stocks in black and through in [indiscernible] Hathaway [ph] here. The reason I did that is I would like to point out that I think Warren Buffet is supposed to bring as much as he does about how much tax I pay. Maybe I do worry about getting his company above water this year, that's 2011. 2010, everybody did well, Cypress is out of the pack. This is one year 2010 beginning in December 31 as the prior year. 2009, Cypress was out of the pack. Semis didn't do all that well. This year Cypress was out of the pack. 2009, semis did well. Cypress beat the semis. NASDAQ did well. 2008 was a horrible year, the best we can say is that we preserved almost 80% of your capital and outperformed in particular the semiconductors where you would have lost half your money. 2007, a great year for us. Warren also had a great year. That's before he got on the tax bandwagon. And okay, so that's the last 5 years. I just wanted to point out. We're very proud of our performance. We look at every year as a new challenge and although Brad is right, I'll agree with them that we're not going to get excited about that kind of return that was the end of business yesterday. The fact is, we're above water. We're making some money this year and not everybody can say that and we're hoping to hang in there for the rest of the year and deliver a return this year and beat the industry.
Three years ago, I ran a one-year long project that was about what I call the solid-state lightbulb. Can I have my solid-state lightbulb please? $2 million auditorium and there's not a plug. Where is it? Conveniently located on the bottom of the podium. All right. So I've got the lightbulb which is disassembled and I just wanted to turn it on, how many watts do you think that is stretched here? 100 watts? How about you? 17? Guy in the back? Tall guy, way back, right there, how many watts? No, you, right. Huh? 5. Okay he wins, 6 watts, and of course it's light-emitting diodes. The reason they look yellow is that underneath the light-emitting diodes are blue LEDs, which are highly efficient, and then they shine through the yellow stuff which is phosphor, they light up the yellow light which then emits yellow light, and when you mix yellow and blue if you're lucky, and you're not really that lucky with this thing, it's kind of a blue blurry light it's not that pretty, but you get something approximating light if you make the phosphor thinner, you get more blue and less yellow. If you make the phosphor thicker you get more yellow and less blue and you can adjust that way. And if they were [indiscernible] which these guys aren't, they put a few red light-emitting diodes in there. They would have had a feedback system to turn the red on and off a little bit and they would snap that white on so you could not tell the difference at looking at this light between looking at this and looking at sunlight.
So I figured, 3 years ago, which I no longer believe, the solid-state lightbulb was the way of the future. And I started working on a projects to figure out how to make this to Cypress. And the idea was to try to make a single-chip lightbulb, solid-state lightbulb, and start producing it with the economies of semiconductors and eventually make a lightbulb for a few dollars. Right today these bulbs are more like $20. And the only way they make money is they convince their senators, I use that word, to vote against normal lightbulbs and make incandescent lightbulb illegal, so we you have to buy their $20 lightbulb.
Now what I really wanted to do, and as shown here, is put a chip on there a PSoC chip. And the PSoC chip would control the LEDs and I'd have the 2 kinds of LEDs. So I have these blue-yellow LEDs, and I've have red LEDs and I control those 2 and balance them off to create perfect white. How would I know it was perfect white? I'd have a color chip on there, now that color chip actually could have been we know how to do this, we've already done it, could've it been integrated in the PSoC so intense color and in a sense reflected light and automatically adjusts to the current in the LEDs to create exactly white light.
Now how do you take LEDs. Which one are run off of batteries, they want to run DC current. How do you take LEDs and take 110 power and turn it into the DC current battery like current that you need. Or that would have been a power chip. And that either could have been a chip we've put into the system or it actually could have been part of the substrate, I'll talk about that in a minute. And then the next thing is if you take about the union electricians, as those lights above you, a long fluorescent light. A single tube costs up $1.57, literally. And once you put in those luminaires, you've got the room. Now let's say you got them in and it's just hanging there and you hook up the power. It costs you another $1,000 to have union electrician run a metal conduit which is something that came out of 1900, up in the end of the ceiling and around and take a break, and hook us to that one and take a break and hook to that one and take a break and 3 days later, hook us to that one, you're looking $1,000. So let's get rid of that. Let's make them obsolete. So I would put an RF chip in there and your light switch over here, you'd have flipped it on you'd have had, seeing and when you flick on the light switch, the little switch would have said your radio turn on the odd numbered level lights or low level lights and bam, those guys are all turned on. Having read with the RF chip what they were supposed to do. That was the vision. It didn't work for 2 reasons. One is, I never could figure out how to make it cheap enough to matter, I just don't believe that people are willing to pay for cool ideas when they cost 2 or 3 or 4x as much as what they've been used to all of their lives; For example, back on the LED versus incandescent light controversy. I read a superior paper written by a professor in Holland and he pointed out that Holland has just made illegal lightbulbs, incandescent lightbulbs. And now we're forced to buy the Phillips solid-state lightbulb. I don't know if you've seen it. It's that thing with that sort of black gray metal on it and then yellow glass on it, it's really, really very ugly. It won some prize of the year in I Triple Magazine [ph]. It's like this one, it only uses fixed watts to produce a bunch of light. But he pointed out by that the argument against incandescent light is that they waste heat. They waste energy of heat not light. He pointed out where he lives, it was cold 10 years -- 10 months out of 12. And that heat went into his house and its helped heat his house and it take away from the heat load of the rest of the house. And he light incandescent lights. And as a human being, he likes the black body radiation curve which is produced by glowing objects, the sun and a lightbulb, and he really didn't need any help from the Dutch government, in particular the Dutch government whose legislators were "convinced" by Phillips that everybody needed to buy $20 lightbulbs in the future, so I gave up for that reason primarily.
The other reason I give up is the white thing in the picture, the substrate. The idea was I'm going to put everything on the chip, let's say that Chip, the white thing it's 1 inch by 1 inch. It's 1 square inch, what we would make in a machine in a factory in, crank them up in the millions just like a lightbulb factory would be those things and you see, you just bring in the power. Actually here's the power right here coming in through this thing, I believe that's a ground wire. And you just bring in the power, make a lightbulb by assembling this and putting some sort of package. Now there's one other problem, if you look at this thing from the side. This is not exactly the job's artistic designer award winner for 2011 and problem is these guys get hot. There are chips that run backwards. They are just like chips, right? You run power into them, part of the power does what you want, like do a computation. The other part of the power heats up the chip. They get hot and turns out light emitting diodes are particularly sensitive to heat. They degrade both in lifetime and in the amount of light they put out, so their efficiency goes down. And they get finicky, pretty low temperature. In other words, our lightbulb gets real hot, it doesn't matter as long you have the fixture that can handle the heat. Whereas these things you can't allow to get very hot or they go weird. So the question was how to put that. Well, the idea was to turn this substrate from -- and this is alum -- oxidized aluminum. So that substrate is made up of aluminum conductee and it's got an anodized or aluminum oxide layer on the top, and those LEDs are therefore insulated by it, but it but is essentially on a piece of metal. And the heat that they create goes into the metal. The metal connects with screws to a bigger piece of metal and then that metal's got dino spines on it and that's the heat gets out of this thing. So my idea was on the back of a piece of silicon, which the world knows how to machine, I would machine a heat sink. I would machine a radiator that would take that heat away from the light emitting diode and at least solve part of the problem. So that was the vision. A lightbulb on a chip, lightbulb.
Okay. This is a cross-section of a chip. This is metal 1, metal 2, which is this in longitude, this in cross section, metal 1. Metal 1, think about this as being something like 50 nanometers. Metal 3, 4, 5, 6, 7, 8, 9, 10. 10 layers of metal, copper. So interconnect in Moore's Law devices has become something we're used but it's really a miracle in terms of the number of connections, the low impedance of the connections. The fact that the connections all fit on top of the transistors. As a matter of fact, a more reasonable paradigm in today's world is not placed in rock which we talked about which means place your circuits down and then wrought them together with wires. In the world today, people often talk about wrought and plates where you put the wires down first and then you put transistors underneath because wires are really with limit your ability to make chip. Problem is if you can't get the wires on the chip, you then go to the other technology which is printed circuit board. This geometry is, let's say, 50 nanometers or 25, 50, let's say 50 nanometers so I can do simple math. This geometry is 500,000 nanometers, 500 microns. So all of a sudden you go from tiny wires that are easy to move back and forth that are very fast and don't just a part of big bulky ugly wires that have been around for decades. And it's kind of nothing in the middle. Second point is you can't put a solid-state lightbulb on there because it would fry itself. This material, there are high temperature resins and more expensive. But FR4, the green stuff you all think about is a print a circuit board. FR 4 is priced with temperatures, so you really can't put heat into it, so that wasn't a possibility. So my thought was, okay, I'll put everything on one chip. Well then you get to another problem and that problem is this. RF chips are typically made in 0.35 micron technology. They typically have a lesser metal which is very thick for inductors. And they're made in foundries principally in Taiwan. Color chips, you can put color on any chip, but you really don't want an expensive chip. So for example, you'd never take an Intel microcontroller or microprocessor which cost something like $0.10 per square millimeter, even $0.15 per square millimeter to manufacturer and waste $0.30 or $0.40 with the silicon put a diode in there to measure color. So you want a cheap technology for that you don't want a lot of metal. the PSoC had that analog and digital and programmable memory and programmable digital interconnect for the PSoC is the zone technology. You can put color on it. You have to modify to put an RF on it, but you can't put power on it because parts just are completely out of it [ph]. Typically those 4 chips are made with different technologies, often in different fabs in different countries even, and therefore the idea you're going to integrate it all together that you're going to make this happen is you realize right away it's not going to happen. And then you look at your next best alternative and you say, what's differentiates me from everybody in China who can slap chips on a printed circuit board. And that's where we ended up. Of course, us the next idea is this, this is the picture I took this morning. This is the card that our VP of Marketing Sales, Christine hands out, the world's smallest type solution. So we're talking about a chip we had which we sold 400 million of them and the chip is in a package called a quad flat pack with no leads. The leads are built into the pack. And if you go to various sizes and you configure. And the smallest one is a 30-ball wafer-level chip-scale package, and there is that. So the point is if you're advanced and if you want to make, keep everything smaller. If you want to get the, whole thing on a 1 by 1 piece of printed circuit board, a silicon printed circuit board, you really want to go to that technology. And there it is, what you do is instead of having leads at the edge events which go to gold wires, which go to pins, which connect to the world that are encapsulated by plastic. This technology right here it wastes a lot area of the chip in this package is probably not a lot bigger than the Arrow, did I mention the arrow on the top of the package X and Y. If you make that arrow squared. That's about how big that chip is. That's an even more ancient package for the leads instead of being inside the package or outside the package.
So who uses it? This is a breakdown of the iPhone 4. Apple has consistently used more advanced manufacturing techniques. There's chips in here or arrow [ph] pieces are the WLCS and CSP chips. This is the cross section where you cut through one so there's a chip, it's upside down and the solder bulb connect it or the chip to a substrate. Now that substrate is important. In turns out that substrate is -- doesn't like to be FR 4. FR 4 plastic PC green material. The reason is that green material contracts and expands a lot. So if I were to heat this thing up, this ball will be forced upward, this bulb will be forced upward and eventually if I heated it up enough, this bulb will actually be torn in half. And the fact is once the chip becomes only 6 millimeters in a linear dimension which is very small, 6 by 6, 36 square millimeters, very much smaller than most chips in the world. The bulbs are already so far apart that, that FR 4 board will rip them apart. Well where do you want to put an outside chip on this. Well, you want to put an upside down chip on silicon. Then the 2 expand and contract together and you don't have any of those forces and you're not limited in how big a chip you can put on the board. So we saw -- we see in the future silicon integrated circuit board, silicon interconnect coming up and that's really, really the point I'm trying to make here. So what does the chips look like at wafer level CFP. This is actually a Cypress Chip. You can see the tag right here that we would use to bond this in a conventional package. You need to route the leads with copper into the center create an array of bulbs on the chip and allow for the chip to be the smallest form factor. That's what Deca Tech does. The technology that does that can do the printed circuit board I talked about also cheaply and that was my vision 2 years ago when I couldn't find a solution. So if we come back here, the name of the game is to make this thing out of silicon. Problem is if you make it out of silicon, that is, for example, if I made a PSoC chip that big and I just put a little PSoC in the corner and then I use the rest of the chip to be interconnect for the other stuff that would cost me $30. And the problem is that piece of, if that were printed circuit board will cost me $0.05. So the penalty for making a PSoC with a big piece of silicon around it is $29.95, which ain't going to work on a lightbulb. So that's the problem we had to solve and how to make that piece of interconnect cheaper. And that's where we partnered with SunPower, a semiconductor wafer, let's talk 8 inch for a minute to a post order so I can just speak magnitude cost $1,000. SunPower makes a full-sized wafer, in their case a 6-inch wafer for under $10 and it's got wires on it. It's got good wires on it. It's got copper wires on it. As a matter fact its wires are more conductive than those wires that need to be right there. So the name of the game is now, going back to the challenge is not to use printed circuit material which isn't good enough, and not to use standard silicon which is $30, which is too expensive but take SunPower's technology which makes wafers 100x cheaper than normal wafers and use it to make the silicon printed circuit board. And by the way, whatever makes the silicon printed circuit board and by the way, whatever the silicon printed circuit board really good, we'll also make weight scale chip level packaging really good. One of the problems with the package you look at it and say how simple, there's no plastic, there's no lead frame. This is it. This is obviously the lightweight answer to the future. It's tiny. As a matter of fact if you use technology like that you could take an iPod and you could make it so small you could actually make a watch of it because everything would be small, you couldn't fit package into the watch because it would be too thick. The problem is that chip's more expensive than a conventional chip. It costs something like 1/4 of a $0.01 per lead. $0.30 divided by 4, $0.075 to package this chip right here conventionally and it cost a lot more than that like a factor 2 or 3 more than that to put these bulbs on it, why? Because these bulbs are not put on with cheap assembly and technology, these bulbs are put on with using fab technology. So even though you go to companies like Amcor and ANC and have them put bulbs on for you, do it yourself, they're not doing it, their usual cost-efficient way. They're, doing it by building their own fab to make their own line to work on wafers. And therefore, they're using very expensive equipment, $4 million machines that run 20 wafers an hour. SunPower's line on other hand is characterized by $1 million machines running 200 wafers an hour. $4 million machines running 20 wafers an hour, $1 million machines running 200 wafers an hour. That's why SunPower's wafers are 200 the magnitude cheaper for silicon wafer out the door than a semiconductor wafer. So the question is, how do you apply that technology and make money with it and that ended up being a partnership between Cypress, an investor and technology donor, SunPower, an investor and technology donor, and our new company, which is the subsidiary company, Deca Tech of Cypress, which we hope will be the next SunPower. And with that, I'll introduce Tim Olson -- oh, Tom Werner came over to give a word then Tim Olson.
Thomas H. Werner
Good morning, Tom Werner, and I'm the CEO of SunPower and perhaps you know our story. It's kind of cool a story actually. The good news, I'm not going to tell you the whole story this morning, but I will tell you that a little bit SunPower because it's pretty relevant to Deca. In 2001, SunPower had the world's highest efficiency cell and the world's highest cost silicon solar cell. And it was a company that was in search of funding and was in fact close to running out of money. And the founder of that company, met T.J., knew T.J. from school at a deli nearby and said, hey I got this high-efficiency solar cell but it cost too much and I need money. T.J. put 2 and 2 together and said well, my company takes costs out of silicon devices. We can help you get cost out and I believe that so much of that I'll personally write a $750,000 check, which he did, the company made payroll and we continue to operate. Cypress then, under T.J.'s leadership, made the investments to take the world's highest efficiency cell, which at the time was a little 4-inch square cell made with photolithography, so very expensive. And made the investment to build a factory, make it in a bigger format and reduce cost rather dramatically. A fairly audacious proposition when we think back in 2002, when solar was a very small industry. And so it took a lot of foresight in terms of market size and in terms of the ability to get cost out and to make this to come to fruition. So Cypress then takes this very small company, 40 people, $4 million in revenue and makes the investment. Fast forward to last year with 6,000 people and $2.2 billion in revenue. So the premise worked. And of the premise was we could take the world type efficiency solar cell and we could take the processes into some degree the people from Cypress and we can invest that in a fab, all simultaneously by the way, so it's fairly, fairly audacious plan. We could invest this into fab and scale this technology at lower cost. And we did that. And we did that on the back of a handful of core people that were able to establish the culture because you're probably aware of this, any analyst who probably track Cypress for a while. There's a little bit of rigor and discipline in this company and that culture that's kind of tighter than the marine corps is built into the way SunPower manufactures products, and therefore the way Deca manufactures products, will manufacture products. So we had processes and people come over. And the processes are incredibly rigorous. The way you order equipment is a stage gate process defined at each step in great detail, the way you characterize processes and you move into production, you don't make subjective decisions of when you do the next investment or when you ramp production, and so we brought those processes over to SunPower and we adopted the culture of Cypress' manufacturing. So we have to make this solar cell for less than $10, 1/10 the cost, and we had to scale this thing. We had to make it bigger. So there was, from day one, a sense that we have incredibly high throughput with low cost capital equipment and because we didn't know how to do it, you had to have idea of rapid cycles of learning. Cycles of learning are essentially running lots getting the result planned a lot, so design experiment. But running it very efficiently, very fast so you could learn very fast. So the core of SunPower was beginning with 5 cycles of learning, which meant that we needed to setup our manufacturing lines to the continuous flow manufacturing lines with low-cost equipment and would run high volume. We run 30,000 wafers per day, per line. We currently have 30 lines. So we build this first fab with T.J.'s help, literally, he helped design the fab the blowers for the cleanroom, the exhaust and I'm sure he's had similar impact on Deca's core manufacturing strategy. As we design that fab, we created a lot of customer equipment. So we didn't go to applied materials or standard equipment manufacturers because it would cost too much. Our entire line manufacturing equipment had the cost what one stepper [ph] cost. So we went to lower cost industries and we customized equipment from a bunch of lower cost industries and we created IP when we did that. And there were dozens of PhDs and manufacturing engineers working on this equipment to create low cost line that was continuous flow, very rapid cycles that could do incredibly high volume. And it succeeded. We went from our first fab with one line, 25 megawatts a year to our capacity this year is a gigawatt. And as I stand here, we make the world's highest efficiency solar cell and we're going to extend that lead as we introduce our minimum 23% cell. So the concept was rapid cycles of learning by having a very fast cycle time in the manufacturing line, low cost capital equipment, it's all connected. And then T.J. had the idea that he explained to you that, hey, we could take some of these concepts and we could apply them to interconnect technology. And along the way, I forgot to mention, I was brought over to SunPower from another startup that Cypress and T.J. had, and along the way, he said, I've got this great guy, Tim Olson, who can do the same thing with interconnect technology but we'll take strengths of Cypress and we'll take the strength of SunPower and we'll bring them together with Tim's leadership. And that's exactly what Deca is. And so Deca is taking investment from SunPower. Its intellectual property license agreement and that is our investment along with some people. And their first fab is built into our campus in the Philippines. It's in the back half of our modco, which is adjacent to our utility infrastructure. So they're in our facility taking ideas from our people and then this continuous flow rapid cycle, customized low-cost equipment concept is at the core of Deca. So we're thrilled to be part of the investment and we think, which we're going to hear from Tim, is you're going to hear a revolutionary approach much like SunPower was that will create a whole new company much like SunPower probably bigger some day, I hope so because we're invested. And at that core is the concept of rapid new product cycles because it built in continuous flow, very rapid cycle times. It's at the core of what they do in phenomenal rigor so that you get better quality and really fast cycle times. And with that, as a foundation for Tim, Tim Olson, founder of Deca.
Thank you, Tom. Glad to be here with those 2 introductions. This is going to be pretty easy and pretty fun. I better keep it short as well since I'm already overtime before I start. Here's what I'll start at. You heard about printing embodied in the cost reduction. T.J. talked about the LED, about what Tom mentioned about solar. And in both of those cases the whole point of it is it brings open a new way to do things, new possibilities around that. And of course, it can change the game. SunPower today is the highest in efficiency, highest technology silicon-based solar company in the world because of this mindset, that's our core. If you look at electronics, perhaps you've heard of [indiscernible] he's one of the leading futurist and author entrepreneur out there. He tells us today for $1,000 in our home, we can buy power that's equivalent to that of a mouse. Not too impressive, but in some ways, put in the other way, you think, wow, I didn't know a mouse had that much power. Because you don't what our computers can do today, right. Well if we look out another decade, that same $1,000 will buy us our own human brain in terms of computational power. If we spin out 2050, all of humanity in our homes for $1,000 computing power. Incredible, incredible, exponential curve, multiple 10x cycle. Of course, what enables this is a curve coming the other way, that's the cost of technology. T.J. talked about those transistor level dimensions and more so. That's kind of cool when you look back at the indicative numbers have let transistors cost or what technology cost. So $1 back when postulated this and as you might recall in '65 he just made a postulation for 1975 and he though it might be over, well it got a lot harder than 1975, obviously. And today, that's 1 nano dollar down at the bottom. So we come into the world lot of nanometers and nanodollars for a transistor. Incredible, incredible progression of technology. Only one challenge and that is that curve is flattening out. And in fact T.J. mentioned it's difficult in integrated systems of the future perhaps to put all this IC technology on that one chip and really make it work effectively.
So that's where we come into play. T.j. Has already explained the picture on the and left and the picture on the right. The picture in the middle is a package. You've all heard of semiconductor package. That happened to be one of the first packages still in existence today called the PDF to plastic package from the 1960s. And what happened there was you took a semiconductor wafer, sliced it up into individual chips, put it on lead frame, wire bond it with gold wire and molded it with black plastic to preserve it. So in the packaging world back then and still today, the chip is a teeny part of what's inside that thing and the reason that package is big, again on the right side of the system-level interconnect is to be able to extend out that package or extend a pitch to be able to interconnect the PC forward technology and most of electronics. So this is the world and scope of electronic interconnect. When you look at the system that we create out of all these technology, of course, we're all familiar what's on the left, but what's simply amazing is what's happened to us what's on the right and what's down here on the table. That is smartphone, for the first time in history, are now the biggest consumer of electronics on the planet overtaking PCs, which for many decades’ computers have been the driver. So half a billion smart phones this year, estimated. Incredible, incredible growth and of course all that power going down inside the smartphone puts tremendous pressure on the electronics interconnection of all those component.
This is study from Global Semiconductor Association and Consumer Electronics published by KPMG, and it talks about when you're out being competitive in this world of smartphones and other advance product, what matters. What matters is getting the product out there in time, meeting the cost, and the ability to adapt to meet the needs of the users. If you boil those down it means speed, cost and flexibility. So if you talk with the top CEOs of all these companies around the world, that's what matters, speed, cost, flexibility. So what if we could essentially create new products in days instead of weeks. What if we could reduce the cost, design iterations for that one feature in the system that's not quite right, and we could reduce the cost of that spend in order of magnitude. And what if we could try new designs in hours versus days or weeks. Well that's what we're about. What's in our name. Where we started, so Deca 10, in Greek, 10x. So our basic company foundation is doing things differently, not just a little bit differently either, by the way order of magnitude differently, 10x differently.
We already heard from Tom. And SunPower really was the inspiration that you could take advanced fab technology, things like diffusions, preservations [ph], inorganic and organic preservations [ph], spattered metals, electroplated metal and do them in a lower cost, high-speed pro line like environment versus this traditional wafer fab-based SaaS-based world that we live in today in creating the semiconductor chip. So SunPower made it happen. There's a picture of inside of SunPower in the right, it doesn't look like your typical fab. And what was really quite interesting is we dug into it with SunPower was there was strong mapping between the SunPower process on the left and what we needed at Deca on the right. First for wafer level chips got packaging and second a very exciting additional products T.J. has already alluded to which are silicon substrates, silicon interposers of the future. So we're uniquely positioned to have a really incredible road map that we think will transform how people put stuff together inside these advanced electronics.
Now it looks simple on the drawing and in fact, when you make that map, there's an order of magnitude down in dimension, and of course, this problem too, SunPower solar cells are 5-inch square coming from a 6-inch semiconductor ingots of silicon. We're dealing with 300-millimeter wafers and 200 millimeters, so that to scale up the technology and scale down the line width the dimension, so there was a bit of challenge involved in that.
So back to this picture. PDIP, well excuse me, before I go there. Our company goes across this space relative to interconnect. So, yes, we're a packaging company, but we're much more than a packaging company. We're backing up into the chip space and we're moving out into the systems space. So we believe Deca will create a new class of company that's all about enabling advanced electronic interconnect inside products in the future. So back to the PDIP, it's going away. Back of a PDIP looks like that today that's our PMEC, power management chip for smartphone and what you see is the chip on the top, flip chip attached to a substrate with the big bulbs on the bottom attached to the motherboard. So there's no wires, there's no mold compound. There is, excuse, flip chip in the package, but eventually, packaging transitions have occurred and this is mostly what's going on the smartphone today because of its size largely. And if you look at the system level that's a structure called POP, package on package, it's in virtually all smartphones. And the top series of chips you see are -- the top several chips all stacked into one -- right there, those are memory chips. So the bottom of that stack is DDR2 for the memory to support applications processor. The top part of the stack is NAND, nonvolatile memory to support programming and content. So that's interconnected to that chip down below, right below the bottom of the cursor now, right on the cursor now, that's the application processor on the phone. That drives the incredible graphics that we see today and everything else and it uses those upper connections to go up into the memory package. That whole assembly called a 3D package or a POP, package on package, is then attached to the phone motherboard. So that by itself is an incredible system that used to be about 5x larger than that, just a few years ago in terms of area in the phone, now it's all condensed to that very small space. But if you look inside, one of -- for instance a small tablet, this is that chip we just looked up. Here's a thin cross section, a little bit bigger view to see. So there it is. It's just one piece of this whole system still, it's just the application processor piece of the memory. Incredible interconnect technology going on in that piece but you still have all the RF functions all the modem functions. Everything else, so the system-level problem just continues to expand.
So again, T.J. explained what a wafer level CSP is. We performed this service on its native completed silicon wafer. So our starting position is that semiconductor wafer coming from the Cypress fab, coming from another IT company's fabs, coming from ESE, UMC companies like that and when we're done, we're selling individual devices out of that wafer, and that whole process of completing the interconnect, all packaging in this case is done in that wafers form. This is actually an RF chip device. Here in cross section you can see here's a chip via down to the interconnect point on the chip, on pad. We have a couple of areas with dielectric. We have a pattern, interconnect patterning capability with the buildup structure of the solder bulb. So the parts build like this, we then flip it over and we attach it directly to the motherboard of the electronic appliance.
It's a big market. Today its $1 billion and 5 years from now it'll be $2 billion. This is just the wafer level packaging market. It doesn't talk about what's coming next, which is in order of magnitude, bigger than this with the same technology.
There's some important IT technologies that are already in wafer level CSP. The top 2 we're not going to talk about, the bottom 2, ARC connectivity, that's the GPS wireless LAN, Bluetooth, FM radio chip that we all have in our cell phones these days, if you don't have a GPS functionality it's your loss. That's light green most of the big players, Broadcom, Cambridge Silicon Radio, et cetera. The blue, the analog players, we all know who the big analog guys in the world are, those are our customers. That's who we're targeting because they're the high-end consumers today of wafer level CSP. And we have 6 customers engaged, we have one customer qualified and ready start production. Pending PCN, OEM, TCN acceptance, the product change noticed. So we are moving down the path. T.J. already showed some of these, but here's the chip T.J. showed close-up, here's a USB chip, here's a CapSense chip, so lots of different technologies going on in CSP.
So back to speed, cost and flexibility. Those things that matter most. Speed. The red bar is 17 days, the blue bar is 3 days. The industry today we have a hot new IC. You want to get it over and get it to Samsung for that phone spin they're going to do in 2 weeks, you can't do it today if you're an IT guy because to get it to this wafer level packaging back end takes 17 days just putting the ball done so to speak, [indiscernible] all traces involve, tests, [indiscernible] it ends up being 3 to 4 weeks from the time to put your IC out of the fab till you can get into the OEM. Big problem for new product introduction. The blue bar incredibly powerful for competitiveness in ICs in winning design, OEMs. AL, the Autoline, we didn't talk about that much, but Cypress kind of invented the Autoline concept. T.J. invented the Autoline concept. It's been very successful at Cypress in the world of packaging. SunPower's foundation for how they manufacture was based on the Cypress' Autoline, so they have a fab that basically runs like a pro line. That's what we do at Deca. And that's why that blue bar is 3 days versus 17 days, it's not that we're running faster, and so we've changed fundamentally how products move through the factory. It's a radical change. We don't believe any other fab in the world runs that way to the best of our knowledge. So it's a really different way of thinking.
From a cost perspective, we already looked at Moore's law. Packaging, the red line here has not kept up with Moore's Law, it has lagged. Our job is to disrupt that lagging curve and bring a new kind of cost performance capability in the industry. T.J. highlighted on the silicon substrate. You have to order of magnitude adjustment in cost to be able to really have an impact. So just we're not quoting specifics today, but we have a very disruptive cost structure and ability to go, we believe, transform this industry.
Flexibility. We talked about earlier design the manufacturing in under 60 minutes, that's incredible. That's a -- it's a week today at least to go from a design to something that you could hold in your hand. So in addition to the wafer moving to our factory fabs, we also have some incredible lithography technology and things that go around a lot to enable that cycle. So for doing design spins and doing system-level modeling, it's a very powerful capability. We didn't want to show you how we do it because that's a little special. So this is a picture of my daughter. She was 11, her name is Maya. She got second in the United States doing that dance in Orlando, Florida a few years ago. She's a little more flexible than I am.
Our headquarters is in Tempe, Arizona. Our factory is in SunPower, as Tom mentioned, that's photo of the building we're in. We've come a long way. We started this in 2009, put the team together in late '09. Really got serious in Q1 '10 and there's all the things that have happened. We did our engineering qualification in Q2 of '11. And now we're all complete and ready to roll.
We did that because we have an incredible base of talent. First we have Cypress supporting us and SunPower supporting us, and you've heard about that. What an amazing foundation we're growing from and we have all the talent helping us from those companies and from other industry sources, so incredible team, we currently have 67 people. We're getting ready by the end of Q2 of '12 to move to 7 by 24 production supporting multiple customers.
Just one a quick word on me. I've been in the interconnect world at Amcor. I lead the research and development there for a number of years. Before that I was in an equivalent world to help bring to market this technology called strip test, which was a pioneering breakthrough and is used industry by many folks today. And back early in life I helped the team in Motorola create that black Startech phone that perhaps you all remember. I think it's still one of the finest cellphones that ever existed. Inside that thing we went from an RF module that was the size of a quarter to an integrated IC with some special interconnect packaging technology that was the size of a pencil eraser. So we pioneered that with the IC technology and the interconnect technology way back to enable the Startech essentially. It was a critical development inside Motorola to enable that.
So first revenue. We're planning Q1. We really pending OEM acceptance of our PCM as I mentioned. Break even, early '13. $10 million quarter Q4 '13, we should certainly beat that. We're, put some numbers up here but we've got a very bright potential to move a lot quicker than this. Cypress has invested $35 million so far. The current plan shows Cypress investigating $86 million. SunPower has also made cash investment, as well as our IP investment in the project and the employees are also investors, that's part of the Cypress startup model. We expect very strong growth as we launch later next year, second half of '12. Our model is 40% to 45% GM and PBT at 20% to 25%. If you look at folks who make wafer level chip scale packaging today, TSMC, the leading foundry, the green line. That's their financial performance. The brown lines are the SATS providers, ASE, Amkor, SPEL, companies like that. This is their performance over the past several years. You'll notice from those numbers I quoted you on that last page, we're a foundry. In a simple word, we're a foundry, we're not a assembly and test company. So we're a foundry, but we believe we're kind of creating a new class of things as we move forward as we mentioned.
We don't like to predict it, we'd like to create it. So SunPower has given us exclusive license to their IP in our domain. And someone asked me earlier in the back, "Well, can other people go to other solar companies and do what you're doing?" And the answer is no. They can go to other solar companies, but no other solar company on the planet does what SunPower does. Most solar cells have screen-printed lines and things on the front that you'll see, including the ones on top of the stop light back there. SunPower cells are actually semiconductor chips built in a fab. Almost -- well, all the rest of silicon-based solar cells are very low technology screen-printed collection of the energy from the sun. So we think we have a very, very special foundation that no one else can easily copy. We have unique equipment and processes. Tom Werner talked of the pain and the work that they've put into that, that we've leveraged. Our operational approach, again coming from SunPower and Cypress. We've got Cypress as a parent with incredible access to IC technology. Got their backing. And of course, the 10x mindset, is perhaps the most important.
From a value perspective, a couple of years from now, we're afraid that folks will want to flatter us by trying to imitate us. So we're working very hard to not let that happen. We're putting a lot of energy in protecting our IP: trade secrets, patents, exclusivity. The first 2 are pretty obvious. The last one we, like SunPower, put a tremendous amount of work early on with -- working with SunPower suppliers, as well as new suppliers to have multiyear exclusivity. And in most cases, in all critical cases, we have 5 years minimum, where people cannot buy our equipment or they cannot buy something substantially similar to our equipment. So we have a very strong position with our supply base to prevent our competitors from quickly emulating us. And in some cases, we have indefinite coverage.
So finally, what are we about? We're about the speed, unprecedented speed. We're about value. We're about flexibility. We're starting out with wafer level chip scale packaging. It's a known market. We went with a proven structure. The first 3 OEMs and the first 6 IC companies we have approached, they all said yes. We like what you're doing, we'll qualify that. So basically we have honed in on what's accepted today, so we can quickly get the company up and running. But what's really, really exciting for the future is what comes next, over the next year or two. So with that, thank you.
Brad W. Buss
Thanks, Tim. Keep that on for a second. I mean, so yes, we thought it was probably a good time to take a break. There's a lot of fresh concepts and ideas. So if you want to do basic Q&A, we'll take that right now.
T. J. Rodgers
So Tim, stay there.
Brad W. Buss
If you ask a question, I need you to go into the mic.
Unknown Analyst -
Tim, I just want to make sure I understand your comment, you're not a T&A company but a foundry. So the company you're working with, are you fabricating their chips and then do the T&A as well or you're taking wafer from another foundry into...
T&A, I like that. I've never been known as a T&A factory. Sounds interesting. You mean a SATS supplier? That's semiconductor assembly and test. So the semiconductor assembly and test providers today, like ASE and Amkor and STATS ChipPAC, they do wafer level chip scale packaging. So they also take a Cypress wafer or another semiconductor company's wafer and they perform the interconnect technology, the building up of the package on the wafer, like we do. So yes, in that space, yes, we are directly competing the same, same technology. TSMC does that same thing. They've gone from a fab and pushed it backwards. So they are promoting that technology. So we are not building wafers. In that case, we're taking wafers from customers, building up the technology, completing testing in some cases, and then shipping to OEMs. Now in the future product, the silicon substrate interposers that T.J. referred to, we will be supplying product, okay, but not competing with our customers, enabling our customers with our future product. Does that answer your question?
Unknown Analyst -
Two questions. Do you see this as eventually something that you would license out to other companies? And then can you talk about any signs of competition or competitive technology that's on the horizon.
Sure. From a license perspective, we don't desire to do so. Sometimes folks approach you with a very interesting business proposition to say, can I have a license so I'll give you something of value, like a 5-year supply agreement where I get a bunch of your business? Sure, we'll sit and talk about that, but it's not our strategy to license people. On the second question again?
Unknown Analyst -
Any sort of competitive packaging technology out there?
So on a wafer level CSP space, the folks who do it today use wafer fab equipment, like T.J. explained. And they are competitors, but the way they do it is a very different way. They use expensive machines doing low wafer counts, high cost, low in flexibility, long cycle times. So they're a competitor, but in the end product not in the method. In terms of our method, is anybody thinking like we're thinking or attack -- I don't know. I don't know of anyone. I believe based on experience that at some point in the future people will want to learn what we're doing. So they may try to emulate us, but not at the current time. Any other...
Brad W. Buss
Any other quick questions?
T. J. Rodgers
We'll have a Q&A for the rest of the stuff at the end. We just want to focus on the DecaTech stuff here. So if you want to stretch...
Unknown Analyst -
Maybe just following up on that -- can you just talk about the patent protection you have? Whether it's the process or whether the equipment that's used, whether, do you talk about exclusivity. You started this business in '10, kind of, if it's a 5-year horizon, and you're already 2 years into it. And then, the different types of equipment if your partner with a certain manufacturer. Are there other guys who do similar equipment? Just kind of, this overall kind of, how comfortable you feel as far as the protection?
Okay. So first off all, it's challenging. As we all know protecting our IP is one of the greatest challenges. On the patent side of it, we're focused on the structures we produce on our future products. The first product is already out there. So there's no patent on the structure or, to apply in that. But our future products, the interposers, some of the other technologies coming. We're focused on structure patents of what we'll produce. We have process patents. We have equipment. These are all pending because we're quite new. And as we know, our patent office takes some time. But so there's a bunch pending. Having said that, Cypress has given us patent coverage. SunPower has given us patent coverage. So we have coverage coming up from that regard as well. So that's the patent story. On the exclusivity, the suppliers we deal with are folks you would not suspect. They're not semiconductor companies. In fact, the reason you can negotiate a 5-year exclusivity quite often, is that we went and knocked on their door, they didn't know what a wafer was. We said, "Hello, we're Deca and this is a semiconductor wafer." And they kind of looked like that. And so we don't go to semiconductor companies to build our equipment because they want to charge $4 million, and machines that go slow. So like SunPower, in several cases, we use SunPower's equipment suppliers in our factory. But in other cases we don't. Where we didn't do that, I would say it's very difficult to find out who they are and where they come from, and we've gone to extreme measures. When you come visit our factory someday, first we're not going to let you inside, but we'll let you look through the windows. And when you look through the windows, everything looks the same. All the same color and there is no name plates on anything. And even if you were to sneak inside and sneak around the back side, even the voltage panel on the back, it would not have a country of origin or a supplier's name. So our specs on keeping everything secret are very intense.
Unknown Analyst -
You mentioned on one of the slides you're targeting RF and power guys or analog guys, have you determined any sort of a parasitic implications of working with those types of things and obviously in RF you're dealing with high-frequency and in power you dealing with obviously power dissipation within a very small space. So can you talk about how this technology is able to take care of some of those?
They love it. The reason they love it is flexibility. Remember we talked about design to manufacture in 60 minutes. If I've got an RF circuit with EMI and crosstalk and issues, what a beautiful technology to try 8 designs all at the same time. See how they perform, first, at a chip level, put them on a board, try them on the board. We have unprecedented capability to help both the analog guys for power or the RF guys for RF performance to optimize the design of that layer of interconnect with our technology. So they are very enthused. They like that.
T. J. Rodgers
Let me just give a little background on the RF, I got it. So RF, think about 2.4 gigahertz, which is a very common band. Electricity is waves, at that frequency it's more like light than it is like electricity. And therefore, the lines you have aren't really conductors like wires, they're wave guides. So you for example would make a line and then you put a line above and below it. That will ground it. So you'd have a transmission line, et cetera. So there's -- it's not really black art, but it's much more artistic than a chip layout. And a lot of times tuning in, in that last design to get the last 2 dB out of your radio is empirical. And what he said is, we go -- if you remember earlier, we go from your concept to a -- I'll leave that alone. We go from concept to production in 60 minutes, the computer does it. Therefore, if you want to take 8 hours instead of 60 minutes, you can go from concept to 8 different layouts with 8 different artistics, renditions of your RF chip. You can put them all on one wafer and you can test them all and see which one's faster and then take the winner, and then call it tile #4. So that kind of empirical tweakability, the concept that you can turn in days allows people to redo their designs much more often, and we call that cycles of learning. In today's world, things are very hard for make. And therefore, no matter how good you are, the more cycles of learning you can pile up before the deadline hits. Christmas is going to be Christmas, cellphones are going to get bought. The only question is, is your chip going to be in the cellphone or not. That's what they provide. On the power side, when you heat up a PC board, the first thing that melts is the plastic. When you heat up silicon, you can heat it up till it's literally red hot. It doesn't matter, the first thing that melts is the wires. So in terms of ruggedness and temperature capability, silicon is a much better substance than plastic. It's a problem. Everybody's always known that. Problem is, it's always been a thousand times more expensive and you couldn't afford to use it.
Unknown Analyst -
Tim, can you walk us through the revenue model? For instance what exactly the services that you're going to be providing? How much are you going to be charging per unit level?
Brad W. Buss
The answer is no.
T. J. Rodgers
Brad W. Buss
It's on the slide.
T. J. Rodgers
Let me tell you the -- let me answer that. How do you get in the door? Okay. The vision is grander than the first product. The first product is, I'm going to make CSPs like your CSPs. The other guys that make CSP. I'm going to make them in 3 days, you're going to make them in 3 weeks. I win for the cycles of learning argument I just gave. And that's very critical, okay? What am I going to charge you? Am I going to charge you of my cost plus some margin? I don't think so. I think I'm going to charge what I can get away with, and what I can get away with, with medium-sized companies. For example, Cypress is a customer. I can get away with their price plus a small premium for speed to market. What can I get away if I'm selling to some monster company? I have to give you their price, maybe minus a little bit, but certainly not the differential between my cost and their cost. So the name of the game is to wedge in to an existing market made -- with a product made radically different way. But then, you have to get some money and get this thing sort of cash flow positive, right? As I cross the $80 million mark, I'd like that to happen. And then we'll start talking about those new visionary product. We're not -- we're already working on them, but we'll start bringing those up later. So the model is, to go in and use our 10x speed capability primarily to wedge into the market, which I wouldn't call it saturated, it's certainly not well served. It's extremely frustrating. It's more expensive as I said than normal packaging and become a player there and make some money on our way to the grander vision.
Tim, If you look at every one of the customers you're dealing with, the speed and flexibility, their jaws drop, then you toss the cost, they get really giddy. Then we've actually had a few guys who are like, hey we'd might invest, right? So I mean there's a huge value proposition for them, and I think the economics are going to evolve overtime and as you get to the newest next gen of products beyond the wafer level chip scale stuff, the value creation is even greater. So I mean, I think it's going to evolve over time.
Brad W. Buss
We're going to cut it there. We can pick it up more at the end and then we're going to bring Muhammad Ali up for, in the ring.
I was going to say Brad, this week Joe Frazier died. So I don't want to be construed as Muhammad Ali necessarily, since they weren't the best of friends. We should respect him this week.
Okay. I thought I'd start off by looking at the transformation of the SRAM business. We're going to take a peek back half a decade or so and then give you a peek forward. And look at it from a few different angles. And by the way, I will mention also that is not Joe Frazier, that's Sonny Liston.
If you look back to the middle of the last decade, 2005, we were well on our way to becoming the #1 supplier of SRAMs. We were #2 at the time. At that point in time we were making our own wafers in our own fabs, 100%. It served us well. But at the same time, we weren't able to react dynamically enough to market fluctuations. And in market demand decreases, we couldn't load up our fabs all the way. So we suffered market volatility and gross margin problems. So it wasn't uncommon for us to run our memory business in the low 30% gross margin range. At the time, we were still doing a lot of technology, 90 nanometers. We were a close second on the technology front. We were running state of the art QDRIIs at 250 megahertz. The random transaction rate, that's a very important metric in the networking space for our customers, we were able to do a 333 million transactions per second, and I'll talk more about that in a minute. We were shipping 4 megabits per unit. So we're delivering a lot of content. At an ASP, above the company average, $1.73, not bad. But a $6 ASP for the product that's growing the fastest, our synchronous products. So that's where we were half a decade ago. We did continuous commitment, continuous investment, took over the #1 position in 2009. At that point, we adopted a flex fab strategy. You've heard a lot about that. That's been extremely successful and it's going to be important, especially not just to take us where we are now, but for the next couple of years and I'll explain why. At the same time, we're inching closer to our 50% gross margin business and we took over the leadership in technology. First in 65 nanometers, first in 550 megahertz I/O, which translated to a doubling of the random transaction rate. And at the same time we were delivering more content to our customers, 6 megabits per unit and the customers pay for that value, $3.21 ASP and almost a doubling in the sync ASP.
While the story hasn't really changed that much. That's a 4-year look at how we progressed. If I go and look another 4 years forward and look out to the end of 2013, we'll still be #1, although the gap will be bigger. We're gaining market share every year, and I'll show you the trend. Flex strategy is working very well for us. It let us keep to our fabs efficient, keeps them loaded and allows us to move product around from factory to factory to service our customers when they need it. And now our gross margin's inching closer to that 60% level. Brad mentioned the 60% corporate level, my personal goal to be at around the company average. We're very close to that right now. And the main message I'm trying to give you is, the business is structured to retain that kind of gross margin for the next several years.
At the same time, lots of technology development, lots of -- the platform development that Brad discussed earlier is not just for PSoC, it's for memories as well. So we are now engaging on 28-nanometer technology. That's going to allow us to do much faster SRAMs than are required, 1.6 gigahertz I/O’s that translates to 3200 mega transactions per second. So now you can see how much we're evolving. In this 4-year time period, we've doubled our capability in terms of random transaction rates. And the next, in this 4-year period we're now going up by a factor of 4 to 5.
And with that, it's going to come larger densities and larger value content, which should be reflected in the ASPs. But the main message I'm trying to give you here is the business is structured and built for profitability, but at the same time we're reinvesting that to continue the development engine.
Market share in the SRAM market. This is a history that goes way back to mid-90s, the same when we were just single-digit player in the market. And we like to analyze data, so I broke this up into different regions. The first one, I call it JAG. That's sort of a sports term for just another guy. So we're establishing ourselves, but -- and that one of several SRAM players just another guy trying to eke out past 10%. We did that. But then we innovated a lot. We innovated Micropower SRAMs. We invented QDR. And that's the healthiest part of our business today. That really changed the game in terms of our growth rate and market share. So that's a 5-year period of time where we now crossed over to 20% barrier. Then steady as she goes for another 3 or 4 years, continue to invest, continue to develop products. Now you see, since we took over the #1 position in 2009, continued success and market share gain, but at an accelerated rate. So we're really separating ourselves from the pack, at the same time there's a market consolidation going on.
So we're ending up the year close to 40% and the real question is, what's going to happen in the future? I could give you the answer, but you're the analyst, so you will probably be able to extrapolate the points, unless you think we're going to accelerate faster. But I'll tell you that my goal is to be in the mid-40s next year, within one year, okay? Now one of the things that's going to help us is the consolidation, you've heard that. Samsung is exiting the SRAM, sync SRAM market. Now they have previously exited the asynchronous market and now they've sent their notifications to their customers to end their shipments at the end of next year. So this is not a big surprise to us. We knew this was coming, but it's a more formal situation now. So what are the customers looking for now? What are their priorities? Well, it's pretty clear. With the market consolidation, they want a security of supply. That's the #1 issue. And they know Cypress has the muscle to deliver what they need. We've got the broadest portfolio. So that goes along with the first bullet, and pick a company that's continuing to invest, so that it won't be a new issue every year with a new supplier. If we can stick with them with the long haul and that's what we're about and that's why we're continuing to gain share. Now at the same time, that Samsung's announcing their de-focus on SRAMs and getting out of the business, we're doubling down. So we're introducing more commitment to product development, more capacity in our factories and more flex fab capacity. And that's what this press release was about.
So with this profit and with this commitment, we have developed a roadmap and a product investment strategy. So we first start with synchronous SRAMs. These are the RAMs that power the internet, these are in the critical memory subsystems that limit performance on the internet. The current product is QDRII+. We've been shipping that for a while. That's runs at 666 mega transaction rate. The next product is coming out this quarter, that's called QDR Xtreme. That gives us a 50% improvement in the performance. Next product is called QDR4 this is a QDR consortium-based product with a few suppliers, that will come out next year. That massively improves the random transaction rate. And then comes 28-nanometers the year after, and that brings us to almost 5 to 6x improvement in random transaction rate. So a massive investment in Synchronous SRAM. But the story doesn't end at Synchronous SRAMs. We're committed to develop and invest in all our SRAMs.
The async SRAMs won't be left behind. We currently have the broadest portfolio ranging from very old, very legacy 64K SRAMs that we made 15 years ago, all the way up to the biggest density in the world, 64 meg. These are fast Asynchronous SRAM memories or micropower memories, super low-power memories. And those are in the 90 nanometers state-of-the-art async technology now. But we're going to leverage what we've -- the technology development we've done at sync and apply it to Asynchronous SRAMs. So we're going to have a complete portfolio refresh on the asynchronous product line of 65 nanometers. And these will have architectural improvements for soft error rate to have a perfect SRAM with error correction capabilities, so that customers don't even see any SER defects. So investment in sync, investment in async, but no SRAM is left behind. We have NVSRAM, nonvolatile SRAM. And this is a faster growing smaller business unit of mine. Nonvolatile SRAMs right now are both serial and parallel and go up to 8 megs. We're going to introduce a whole new family on 65-nanometer nonvolatile technology. And that's going to happen over the next year or 2. And this technology is actually going to be the driver for future PSoC generations. So there's a lot of company synergy. The message is, no SRAM is left behind on the technology development.
Okay. So then you think about the future. The future of Synchronous SRAMs, which is really powering our growth. Some of you call me and ask me about DRAM infiltration and what about competing architectures? And I'm going to talk about that. First, why is synchronous going to be so healthy over the next decade? Here's one thing in our favor, video, Internet content. So video content now represents close to 90% of the traffic on the Internet. I saw an article last couple of weeks ago that said Netflix consumes something like 60% of the bandwidth on any given night. And actually -- who have Netflix now? Who uses -- keep your hands up. Who uses the mail-in disk, DVDs instead of streaming it online?
Both, right. So a lot of hands went down. So the fact is, only about 10% or 15% of the users actually stream the content online. And you could see that trend increasing massively. So even with that low adoption rate, 60% of the Internet bandwidth is clogged up on peak hours today. This is coming from Cisco, Visual Network Index Report. And this shows petabytes per month trend on the Internet, doubling every couple of years and the main message here is video, video, video, Internet, video, file sharing, which is heavily dominated by video, video calling, video TV and the gaming has a high video content. So video is clogging the internet. That means you need to do lots of random transactions, lots of managing that video data stream. So what does that mean to our customer? So this is what our customers need to do to keep up with that trend. This is their line rate that they need in their networking systems. And this comes from the Ethernet Technology Summit. So bottom line, the white one is the server market, that's not the market we serve. we're in the networking space. But even that space doubles the performance required every 24 months. In our space, that's doubling every 18 months. So even faster than Moore's Law kind of increase. And whereas it took the most part of a decade to go from 1 gigabit E to 10 gigabit E over the last decade. Over the next decade, we're going to go from 10 to 100, 200 to 400 to 1 terabyte E, all within a span of about 10 years. So that's a challenge. To put these 2 things together. The video dominating the Internet, mobile devices really taking off, more and more cloud computing. And the bandwidth that our customers need, and the implication for the memory subsystem is, we need very fast random transaction rates, the number of independent reads and writes that I can do on the memory in a period of time.
Okay, so how does this relate to sync SRAMs versus DRAM architecture? I wanted to give you a quick example. Let's say our customer has a core router and they want to do one read in a 32-bit wide packet. So their managing data on the Internet, 100 gig E space, so this is like 2012, and they're going to use a DRAM, DDR3. State of the art DDR3 DRAM. In this particular case, to do 100 gig E transmission, they have to 150 million packets per second. That's what it translates to. So that's reading every 6.6 nanoseconds. Okay, the problem though with the DRAM is it can only do a random read every 48 nanoseconds. They got a big problem. So you can't use DRAM. Well, you can. There's a trick you can do. You can bank the data. You can put it into multiple banks of the DRAM data. This is a trick that system architects do. So they'll bank -- they'll do 8 different reads and they'll use 8 different parts of the memory and they'll do it in parallel. So it's inefficient, but they'll do it. But the second problem is, with this device you can only open up 4 banks at the time. So now you've got to use a second DRAM. So now, you've gone from 1 DRAM to 2 DRAMs to do this read. But then you run into a third problem because you want to a 32 wide and these DRAMs can only do 16 wide, so you have to double up again. So now you're looking at 4 DRAMs. There's your packet processor, 1, 2, 3, 4 DRAMs with this banking structure.
Okay. Cost per bit on a DRAM is much lower than an SRAM and 4 to 1 is kind of a okay trade-off, but still looking pretty ugly. But in the real world situation, when you're doing a forwarding lookup in the core router, that's going to require 4 to 8 reads per packet not just 1. You have to multiply this times 4. So now you're looking at 16 to 32 DRAMs to do the operation of 1 QDR4. And that's the advantage we have, cost. Now the cost, the solution cost, favors SRAM. And that's why SRAM is so critical for the next decade in these applications, pin count, a massive of board space and power. So in case you didn't completely understand that example, I'm going to simplify it.
Does anybody remember the Partnership for a Drug-Free America, 1987? This will refresh your memory. These were very effective commercials on TV. They said, this is your brain, then they'd say this is your brain on drugs. See that's a rhetorical question. Any questions? So let's look ahead circa 2012. This is a networking line card. This is the processor, the ASSP or the FPGA, you got one QDR4 with 150 traces. That's your line card with QDR4. That's your line card on DRAM. Any questions?
Okay, let me wrap up. The Memory Products Division is structured correctly with the right products and the right technology to have high gross margins and profitable business. Share of market continually gaining with a special dislocation this year with Samsung exiting. We're best positioned in terms of breadth of portfolio, ability to deliver, and ability to secure supply. Product and technology investments across-the-board on all SRAMs, and QDR is the architecture of choice for the future. Say no to drugs. Thank you.
T. J. Rodgers
To clarify, when he says massive, it doesn't mean we're going to be spending a lot of money. We're reallocating dollars from [indiscernible].
Yes. I mean, we're still be able to keep our OpEx near 20%. So in the grand scheme of things, I'm behaving myself.
T. J. Rodgers
Massive for a memory guy, small if you're Norm.
All right. So my name is Badri, I am the EVP of DCD. I'm going to talk about USB 3.0 driving revenue growth. So please look at agenda, the USB is the most popular connectivity interface to date. The USB 3.0 is here. So 2012 is the coming out party for USB 3.0. Cypress is #1 in USB.
T. J. Rodgers
Badri, why don't you take one minute before you start and tell them about your history at Cypress.
All right. I've been at Cypress for 16 years. I was a technology development engineer first, moved to design for about 8 years and then I moved into business. I was working for Dana. I was running the asynchronous memory product line and then I basically started to manage USB and now in DCD. I'm also responsible for India. I'm the country manager for India. So that's the history.
All right. So the agenda, USB is the most popular connectivity interface to date. The USB 3.0 is here. 2012 is the coming out party for USB 3.0. The #1 in USB, Cypress is #1 in USB. We've got 2 exciting chips that I'll talk about in detail. One is FX3 and the other one is Benicia.
So a brief history on USB. USB was founded in 1996. Essentially USB 1.1 was called -- and now it's called low speed, full speed USB. So primary applications are human interface devices. Then in 2000, USB 2.0 was introduced, about 480 megabits per second was the interface speed and basically personal storage, office automation and camera, the memory stick. And now, USB 3.0 was first introduced in 2008 and slowly the ecosystem is building up. Primary applications will be video, storage and imaging. So 12 megabits per second, 480 megabits per second, 5 gigabits per second. So 10x improvement from USB 2.0. The USB is universal, so universal serial bus, is the most part proliferated connectivity technology in history. 10-billion-plus installed base and 3-billion-plus annual shipment. This includes both discrete as well as embedded USB. So all kinds of applications, computing, consumer, communication, industrial, military, and even aerospace. Increasing functionality. In addition to data, the newer applications of audio, video, battery charging, power and the USB Implementers Forum is constantly innovating. The next step of innovation is for USB to provide a lot more power. So you can drive everything through USB. So it's a concept of 1-chip dock on USB. Almost, the power that we're talking about is close to 100 watts here with the new power delivery class of USB 3.0.
So what's happening on the USB 3 space? So as I said, 2008, the specification was introduced. The first certified host controller came out in 2009. As of today, there are about 275 products certified. So the ecosystem is about to take off. What needs to happen for that is essentially the big microprocessor guys has to basically integrate USB 3.0 in their chipset. And that's going to happen in early 2012. So summer of 2012, USB 3.0 is going to take off. So once the PCs come natively with USB 3.0. A whole new ecosystem will start developing around the PC with USB 3.0. In addition, Microsoft has promised a seamless experience of USB 3.0 with Windows 8. So these 2 will drive USB 3.0 mainstream in 2012.
So Cypress and USB. Basically Cypress means USB. And we've been shipping USB products since 1997, 1996 actually. Full coverage of USB markets. So low speed, full speed, high speed, as well as super speed right now. Super speed and USB 3.0 are the same thing. And we participate in all segments, host, hubs, device and even wireless USB. So as I said, we are #1 in USB market share, and we are actually 1.5, we have shipped 1.5 billion units to date.
So I'm going to talk about 2 chips that we have introduced in USB 3.0 space. One is called FX3. We launched this in April 2011. It's been sampling for the last 6 months. And the other is Benicia. This caters to essentially tablets and cellphones, while FX3 caters to a variety of general purpose applications like industrial cameras, gaming controllers, scanners. So broad customer base and then a few customers here primarily focus on tablets and cellphones. But very high volume.
So FX3. The simplest way to think about this is, it adds USB 3.0 to any system. So essentially, we are talking of a chip that takes a parallel interface, either from a microprocessor or an ASIC or an FPGA or an image sensor and converts it into USB 3.0, industrial cameras, gaming controllers are a good example. And then a typical camera system today, because of the limitation of USB 2.0, there is a lot of compression involved. With USB 3.0, since the pipe is fat, which is 5 gigabits per second, the compression does not need to happen here in the ASIC and the compression can all happen in the CPU, and the CPUs are getting much faster too. So it's not a big load on the CPU. So essentially 10x faster, enables uncompressed data transfer and simplification of the design. So no longer a compression chip is needed here. And we are even trying to integrate things like the ISP into our USB 3.0 chip going forward.
So Benicia is the chip for cellphones and tablets. Here, it is now possible to transfer a high definition movie, which is about 5 gigabytes in less than a minute from your PC to the phone. That's a big deal. It was not possible before. So we expect that to drive a lot of adoption in cellphones and tablets. So essentially 10x faster and synchronizing with the PC, 2x faster battery charging because USB 3.0 supplies 900 milliamps of current versus USB 2.0, which is 500 milliamps and much more responsive user interface because Benicia makes the storage very efficient. So you'll see when you're shifting between applications on your tablets or your smartphones, the response time is much faster.
So the success story so far, we have introduced the product 6 months ago. Like what I said, all metrics are actually ahead of expectations, over 350 customers evaluating, very broad customer base for FX3 and about 20-plus customer base for Benicia, all the big cellphone guys. Over 200 evaluation kits have been shipped to date, $75 million opportunity pipeline. The ASPs are actually 50% to 100% higher than USB 2.0 because of the value that we provide. And I expect a significant ramp in 2012. It is going to be steady for over 5 years and that is basically what we have seen in our previous generation of USB 2 product and USB 3 promises to be the bigger, with all initial indicators ahead of expectation.
So summarizing, the USB is the most popular technology. USB 3.0 is going to be a paradigm shift. It's going to start adoption in 2012. Cypress is the leader in USB. And we have, we are well positioned for USB 3.0 ramp with 2 chips, which is a general purpose USB 3.0 controller FX3, and Benicia, which is, which targets the handsets and tablets. And we are also working on a USB 3.0 hub which is coming very soon and that will add to the ecosystem as well. All right, that's it.
T. J. Rodgers
One point that Badri didn't make clearly, and I want to make now is that, he's a our fourth division manager. He's taking over Data Communication Division. The guts of which is the USB chip. There's other stuff he didn't talk about today. The other thing he's going to do for us, he's been here for, he said 14 years? -- 16 years. He's moving to India. And he's out of here this month, right?
T. J. Rodgers
Next month. You're stalling. And so one of the advantages you have in India of course, is you can get people that went to great schools for, let's say, 1/3 of the wage rate of the U.S. The problem is, you're 12 and 1/2 timezones away in Bangalore and you may lose all of that advantage. As a matter of fact, it's often common that you lose the advantage because you're not synced up with that organization far away, they don't hold your values. Badri has been around a long time. He holds our values. He's part of Marine Corps. And he's going to go run our Indian site for us. So this is our first division that's going to be in another country. The business is going to be run from there. And we're going to have our business and marketing there. Their time zone shift to where our customers are today is less than what it is from here. So this is actually even more important -- we talked about products today, but this to me is even more important for the future of the company. So Badri took over what was Datacom and Dinesh, who ran Datacom is moving to run a second PSoC group. PSOC is getting big enough that we can't handle having just one PSoC group. And that's all of our opportunities in particular a failure mode is norm without getting the question, would I like to talk to those thousand customers in the Midwest each of whom wants to buy a thousand units? Or would you like to fly to Korea for the 10 million unit Samsung deal? And you know what that answer always was and the net result is, we haven't been building up our base brick by brick, the way we should have been. So we divided the touch and other user interface businesses. Norm's going to continue to run them. They were a big part of, a majority of what he did anyway. So that's pretty much the same story. So the new division is the base PSoC division where we're going to build up our category B as we called it, or our second tier customer PSoC with our basic design all over them. And Dinesh is going to head that effort for us.
So on that note and introduction from T.J., I'm running the programmable systems division. This division was basically created for us to focus on the broad adoption of PSoC. Our goal is to make PSoC the embedded design platform of choice and what this means is any an embedded system designer, when they think about a design, we want them to start thinking about PSoC and how they will solve their problem using PSoC and the tools vested around PSoC. Our goal is to drive large customer adoption through distribution. We aim and strive to make the software and our tools very easy to use and very easy to learn. And our focus in development for this group is going to be primarily on programmable analog and programmable digital. We are enabling an ecosystem around the PSoC platform. This ecosystem allows partners and customers to quickly design their systems and put those systems into production, and I'll talk briefly about how we are planning to do some of those things as we go forward. And this group also encompasses our automotive effort, which will now be focused and be driven through this particular division. So we are focused on driving and proliferating our automotive PSoC wins, some of it you are seeing in cars and other things over here, in human machine interface, HMI and in body electronics.
So to start off with -- this is a picture that most of you have seen. This is how we have been positioning our new introduction PSoC 3 and PSoC 5 into the marketplace. We walk in and convince our customers and our partners and our distributors that MCUs cannot do certain functions, whereas when you use PSoC, you actually can. And a portion of this presentation will be to explain to you how that comes about and what is the value that we are bringing to our customers and the value that our customers end up receiving in the products that we are offering them.
So to start off, there have been over 1 billion PSoC sold. So this is not something that we are new to. We've been making PSoC for quite a while at this point in time. So we started up in 2000. We've run through selling about 2 million units in 2002. And we have gone over 1 billion units in 2011, and this shows you what the ramp has looked like.
PSoC 3 and PSoC 5 are our general-purpose devices that we introduced into the market. We are seeing a very strong uptake off of those products into the marketplace and the reason is because the PSoC creator software is unique, a very differentiated in the marketplace. And it's seen about 15,000 downloads to date. We've also introduced a new version of our creator software and I'll briefly mention that in the next few pages.
Development kits. We've sold more than 9,000 development kits to our customers. And we have over 600 design wins that we are in pursuit of or have won at this point in time. If you compare our sales funnel with how we perform with respect to PSoC 1, we are significantly ahead both on the sales funnel as well as on our design win ramp and the community of PSoC users and PSoC lovers as gauged by the developer community that we have put together has grown to over 40,000 users.
So when we introduced PSoC 3 and PSoC 5 we had a massive TAM expansion. The reason we had massive TAM expansion is because PSoC 3 and PSoC 5 now allow us to compete in the high end of the 8-bit microcontroller market. Included in that market is also a significant portion of the analog space, and the PLD space as well. In addition, PSoC 3 and PSoC 5 allows us to compete in the low end of the 16-bit space with our 8-bit microcontroller. And with PSoC 5, it allows us to compete in the beginning portions of the 32-bit microcontroller and covers the entire 16-bit range. So our overall SAM when we introduced PSoC 3 and PSoC 5 basically got expanded from $1.6 billion all the way to $12.7 billion going to $15 billion by 2012. So we have a very gigantic market which we are beginning to address with the partners we have brought to the table.
So the key differentiation when you compare PSoC 1 versus PSoC 3 and PSoC 5, is our ability to add more microcomputing, microprocessor computing power. So we've gone from 30 MIPS -- we've gone from 4 MIPS all the way to 30 MIPS on PSoC 3 and up to 100 MIPS on PSoC 5. In addition to that, we have significant analog devices that we have introduced into our offerings in PSoC 3 and PSoC 5. You can actually do an 8-bit kilo samples at 384-kilo bits -- kilo samples per second, all the way up to 1,000 kilo samples per second on PSoC 5. And the programmable logic density that we have introduced where you can put in your customized code into our chip has gone from the 2,000 gates that we offered with PSoC 1 to 20,000 gates which is what we offer in PSoC 3 and PSoC 5. These 2, the analog density and performance and the programmable logic density that we have included are the things that our customers are using a lot more and allowing our products to differentiate themselves in the marketplace with their system.
So the high-performance analog, the way we have constructed our PSoC architecture looks like this. So we have what we call prebuilt components inside of PSoC. What this means is you can have a ADC, which is a DelSig device. You can have an ADC, which a SAR device, which are all analog to digital converters. Typical analog companies in the marketplace will actually charge you for an ADC if you want to go buy it as an individual chip into the marketplace. We actually include that chip as part of a component inside the PSoC infrastructure. So in our software, the user can come in and say, I need ADC SAR 1 and you can drag and drop that into your design and then start using that SAR ADC as if it were an independent chip that were sitting essentially on your board and the PSoC microcontroller was communicating with that ADC to get data from it or you could walk in say, I have IDAC. I want an 8-bit IDAC. You can drag-and-drop that in. Now you have the ability to put data out and the ability to actually get data in into your microcontroller for doing various kinds of computation. So what we've done inside our software is we've prebuilt analog components and we keep building more and more analog components and offer it as part of our analog component library to our customers. So we keep adding more and more of these components every quarter and that improves the usability of the tools. It improves the usability of what the customers want from us and it allows them to become very, very addictive to our software that we offer them. Because we are abstracting the capabilities of our device, and we're giving them more complex things to actually to deal with.
You take those components, which are prebuilt, you combine it with a routing fabric, which has significant number of connections, and you can now mix and match a whole bunch of analog components together, which gives you much higher degree of functionality. This basically means that you have thousands of unique analog circuits for which our analog competitors would charge you a specific amount of money for a discrete circuit. They're all built into PSoC, and the user has the ability to pick and choose those analog circuits that are necessary for them for their system design.
So let me walk you through an example here. So if you look at a standard MCU offering from one of our competitors, this black box is what you will see inside any standard MCU. What would it be? It will be some kind of microcontroller, in this case, an ARM Cortex-M3 for instance. There'd be a bunch of GPIOs, or general purpose I/Os. There'd be 1 SAR ADC, which you then have an analog multiplexer to take a bunch of inputs in, have a PWM or pulse width modulator and you'd have an I2C master interface to actually communicate with the external world. But within PSoC, what you get is this microcontroller built in and you get all these other devices, which are as, I told you before, components which you can connect up together and connect up to the main component and make this entire device into a power supervisor and a power thermal manager for you. So you don't have to have discrete components that you need to put onto your board, you actually just include them inside the PSoC environment for you. So the value that we bring to the table is what you can do with an MCU and even more. And this is what our customers are beginning to realize and some of these components are fairly expensive components that they would typically end up having as discrete components on their board and that's where we're seeing the value of PSoC to our customers.
So PSoC 3. Because of this architecture is opening new markets for us, markets that we have played in, but in limited capacity. So for instance, some of the design wins that we are going after are power and thermal controls in servers. So there are lots and lots of servers being shipped into the marketplace. We are able to walk in there and basically design a power manager and a thermal manager for them. We're looking at door locks, in which we're able to combine the CapSense or the touch capability along with what we can do with solenoid and being able to move those objects, as well as do communication to get into the digital door lock market.
4G Wi-Fi router. We are basically able to provide customers USB, CapSense and battery charging capability. This is another interesting thing where we have -- a company, which is basically making iPad test and measurement devices using our device. So it converts your iPad into a very powerful test and measurement device. We have top fitness equipment makers, which allow you to run a motor or drive a conveyor belt that allows you to drive the fitness equipment. We also have wind turbine, where we're looking at multiple sensor inputs coming in from the turbine and then figuring out how the motor gets charged or how to discharge a motor and basically convert that energy in a battery of some kind. So there's a whole new markets that are getting opened for us because of the ability that we're bringing to the table.
We also have lots of PSoC wins in autos. I mean, you're seeing one of our biggest auto wins at this point in time. We're also looking at several infotainment units for German OEMs and their primarily preview buttons. We are also looking at consoles, which is actually the one that you have there. And then we are looking at navigation stacks for the Japanese market. So there's a lot of activity that's showing up with PSoC in the automotive space, and we are proliferating those designs not only to the capacitors, the sensing or the button replacement, but also to entire console replacement. In addition to that, our PSoC 1, which was our junior PSoC, if you might want to call it, is still seeing a lot of design wins, activity continue. We're getting design into microwave ovens, washing machines, e-bike is there. We continue with our leadership, door security, handheld vacuum cleaner with battery control, flash strobe light. There's a plethora of design wins that we're seeing and PSoC 1 momentum is continuing into the marketplace.
So in addition to PSoC 1, 3 and 5, we are working very heavily with our distribution partners. And we wanted to show you what our distribution partners are saying about, what they see with respect to their customers and PSoC. So we have Jeff Eastman, who's the Senior VP of Global Supplier Marketing for Arrow Electronics.
Hi, my name is Jeff Eastman. I am the Senior Vice President of Global Supplier Marketing at Arrow Electronics. Arrow is a worldwide provider of electronic products and services to over 115,000 customers and doing business in 52 countries and 340 different locations. I'm here today to talk to you about Cypress Semiconductor and our relationship with them. I've had the pleasure to work with Cypress for the bulk of my career. They're a very important supplier for us with strategic promise and the future with them looks very promising. They've gone from being primarily an SRAM company to still the leader worldwide in SRAMs, but they're also changing the future with their new PSoC which is Programmable System on a Chip products. They've got a set of these devices, and they are really revolutionizing the way that customers can design and develop products and the way that we, as users, interface with our gadgets, if you will. Everybody knows iPad, iPod, and Cypress has products that enable those kinds of touchscreen technologies and they are doing quite well.
For those of you that like to drink Coca Cola products, be on the lookout. You will start to see a new fountain drink dispenser that Coke is rolling out currently. Cypress is enabling that technology. They got, I believe it's 10 to 15 of their PSoC chips inside that machine that allow it to do some really wonderful things. And Coke thinks this is going to be huge boom for their business.
As we look towards the future with Cypress, we believe that we will winning many, many designs with their PSoC products and this is really fun and exciting for us. We get to work with customers and really change the face of our industry and get involved with some pretty cool products that gets off to the market. One of the things we work hard with Cypress to do is to get to these customers early on, so that we can the design and also help them with time to market. So I think that our future together is very, very bright, and we look to really growing our sales together.
All right. So that was Arrow. I wanted also to give you review a brief view on one of the products that we are showing or sharing with Arrow and our customers and the way we look at the product offering. So this is a unique way to look at product offerings. But on the x-axis, you have UDBs or what you will call programmable digital blocks. So there are 16 programmable digital blocks in these devices, 20 of them and 24 of them, and on the y-axis, there's a 12-bit analog device, this is actually a Delta Sig and a bunch of other analog devices. So the way we look at our parts is on the y-axis, we look at analog programmability. On the x-axis, we look at digital programmability and that's what we're offering to our end customers.Those are the 2 axis on which we end up differentiating our parts, and those are the 2 axis on which our customers believe that they can introduce products into the market much more rapidly and differentiate themselves because they have this flexibility and scalability that's offered by our products. So that's PSoC 3 family.
PSoC 5 family, again, on the same scale, 20 UDBs, 24 UDBs and then of course, we have more UDBs that we can add in to basically give you a much larger number of gates that you could actually end up programming on your device. So this is what the folks at Arrow are seeing, and they're able to position this to their customers and offer it to their customers so that they can start their designs quickly and bring them to market quickly as well.
So in addition to the hardware that we talk about, there's a big component, which is what our customers see us to begin with, which is our software, and in our software, the most important revision that we made is PSoC Creator 2.0. This piece of software basically revolutionizes the way you end up doing designs. It has components built into it. It has the ability to do timing-driven routing, which is one of the key things our customers ask us for on the digital front. It has the ability to do analog viewing and schematic editing. It also allows us to have what is called "component packs". So we have the ability to keep introducing new and new components and allow our customers to incorporate these components into their designs pretty much every 6 weeks. And of course, as we keep introducing new components, the base IP or the way our customers end up looking at the world keeps increasing, which allows them to be a little bit bolder and do more designs, which allow them to differentiate themselves into the marketplace. So one of the key things that's there in all PSoC Creator is third-party IDE from ARM. So PsoC 5 has an ARM Cortex built into it, and we have been working very closely with the folks at ARM to actually enable that technology and enable the integration of the Keil micro version IDE. So to talk to us about how the partnership between ARM and Cypress has been working, I have Simon Segars from ARM, he's the EVP there.
Good morning, everyone. My name is Simon Segars and I'm with ARM. I've been with ARM for over 20 years and spent the last 4 of it based out here in Silicon Valley. My office is in fact just basically across the parking lot there. A few years ago, in fact, before I moved out permanently, I came to a meeting here at Cypress to talk about adoption of ARM technology into the PSoC range. It's great to see that this is now coming through.
So just a little background on ARM. ARM is going to be 21 years old in a couple of weeks' time, and we started out with the company with a view that we couldn't do everything ourselves, we weren't going to manufacture anything. We were going to build a partnership to enable the proliferation of the ARM architecture through as many silicon devices as possible. We could see that there are large closets [ph] of applications where 32-bit microprocessors were going to be required, and if you had a very low-power, very small, very right-sized Silicon-efficient processor, then that could be put in lots and lots and lots of different places. We didn't quite envisage quite how large that was going to be, but yes, we're really pleased with where we are today. ARM is best known for having a huge market share in mobile phones. It's pretty difficult to find a mobile phone these days that doesn't have at least 1 ARM processor in them, and in many cases, there are 4 or 5. But we're also successful in other markets where the processor itself is deeply embedded inside the device and isn't necessarily visible to the person using it.
This drives -- you could see there, we have a very large market share, digital TV, set-top box, quite a respectable market share and growing; and then at the really small low-cost end of the spectrum, the microcontrollers. A few years ago, we could see that there was a migration to 32-bit, a need for higher and higher performance as constraints such as energy control, overall efficiency, safety, security were driving the need for more and more processing. So that was a market that we could see was interesting and a market that we wanted to go and partner with people to help develop with the ARM architecture.
ARM started pretty small. There was only 15 other people when I joined the company. We now have over 2,000 around the world, and I share TJ's concerns about the developing teams along their way. Part of my team is based in Bangalore and absolutely the challenges of dealing with itself in a half-hour time zone are not small. So you have to work on that. You have to work on building that global team. We've done that to make sure we're close to our customers, and again, like Cypress products, we're having a very strong technical content across all our team, particularly in the management chain.
The ARM's company is built on partnership. We rely on our partners to proliferate our technology has been very successful, and last year, over 6 billion chips containing ARM processors were shipped by our partners. And that's growing, and lots of the end markets in which we operate are growing, too.
As I said, it's all about partnership, and our founding CEO basically told everyone we can't do everything ourselves. We have to go and work with companies that have supply chains, to work with companies like Arrow as we just saw to enable design wins, and that is what we've been doing. We've signed over 800 licenses for our technology; over 275 partners now. It's slightly out of date. And you can see that cumulatively, the deployment of ARM processors has been growing enormously, and out in 2020, we predict there'll be over 150 billion chips based on ARM processors that have been built across the entire semiconductor industry.
So that brings us on then to the collaboration with Cypress, and this is something that we're very excited about. What ARM has set out to do is to provide a microprocessor architecture and all the supporting ecosystem around it so that people can design very innovative silicon chips without having to worry about designing a microprocessor. If you look at the PSoC range that exists today and without using ARM processors, then Cypress would have had to develop that themselves, and that is cost. What ARM has tried to do is to enable the microprocessor and the supporting ecosystem, so that companies like Cypress don't have to do it themselves. They can spend their time on all the real value-added technology that goes around the processor.
You saw in the next slide that the processor was just their building block. What's really interesting about the PsoC 5 devices are all the configurable analog and digital components that you use or can choose to differentiate their end product. So every product does not look the same. They can all be different, and that's a key part of what ARM is trying to do, and that's why we really enjoy partnering with Cypress.
In other areas, I think you're going to see in the next presentation an ARM-equipped [indiscernible] is being used in a touchscreen controller. Touchscreens lets you live on a different planet. You have seen it proliferating everywhere. I see my young kids kind of walking up to the TV and poking it and being surprised that it doesn't respond. It's amazing how quickly that has become just the way in which we interface with devices and children just expect that. The other thing that I find interesting here as well is my youngest son had has a few issues. He loves his iPad and runs around pushing and dribbling all over it. When you get moisture on a touchscreen, it basically stops working, so the technology that, as I understand that Cypress has developed to really improve on dealing with background noise, the signal-to-noise ratio, is going to help that, and that's certainly going to help my family. So looking forward to seeing that in our products. And then there's the USB device that again you just saw. That requires a processor. I truly believe it would have been a waste of time for Cypress to develop a processor just for that, so leveraging the ARM technology and all the work that we've done I think enables them to get their device to market more quickly, more cost-efficiently, and help lower the cost of that device going to an end product, and we believe that's good for everyone all around.
So I think there's a huge overlap between what ARM is trying to achieve, what Cypress is trying to achieve to lower the cost of electronics going out to the industry and improve the kind of end devices that consumers like buying. And let's face it, we're all consumers and probably like all buying them.
The Keil tools that Dinesh mentioned is a very important part of that. Keil is a sub-brand of ARM. We have a tools business, and the collaboration there is around ensuring that a user, somebody who wants to use a PsoC 5 device and take the ARM tools, and what the device can do is reflected in those tools so that the user can write their code very quickly without having to go spend lots of time configuring everything and debugging it. So working together, again, we're trying to shorten the design time and shorten the time to market for an end customer device.
So growth is basically which [ph] ARM and Cypress has been collaborating, we look forward to working together in the future. Thanks very much.
Okay. So to kind of summarize, PsoC 3 and PsoC 5 have been the new products that we have introduced into the marketplace. They've increased our TAM very, very significantly. PsoC 1, PsoC 3 and PsoC 5 all are different but they still enable our customers to build their systems and put their products into the market in a quick and timely manner.
On the last part, I wanted to point out that we have a very, very strong and deep IP position with respect to the PSoC family of parts that we have created. We have a total of 529 patents which are either filed or are pending, and these patents include our technology in building PsoC 3, PsoC 5, the programmable analog part, the programmable digital part, as well as the touch interfaces, which Norm will be talking about next. Typically, we're ranging at about -- filing about 50 patents per year, and we hope that that number will actually keep becoming larger and larger as we go forward. So we have a very, very strong IP position, and for all those who have known TJ for a long while, this is one of the key tenets for the company, which is we make sure all our intellectual property is filed and protected over a period of time.
So in summary, we are seeing PsoC 1 continuing to win. We are design win -- replicating the designs that we're seeing and pushing it into more and more customers. 3 and 5 are the foundations for our businesses going forward. It's one of the reasons I'm sort of running this division, which is completely and only focused on building up a large base of customers and making sure that we can use our partners and our customers and push the technology that we have built over a much longer period of time to our customers and enable our customers to be successful by introducing their products into the marketplace. And that's all I had, and I think Norm is next.
Norman P. Taffe
Hi, everybody. Good morning. It's almost good afternoon. We're almost done. I am the last speaker. As TJ and Brad pointed out, I do have a different role, a really narrowed role. The reality is, in one of the things we did in PSoC, we had to tap them [ph], turn it into TrueTouch, and frankly, it was so successful, it's built its own division. Even if we separate it aside, this division will be the second largest in the company, and actually with growth trends [ph] may become the largest within the year. So that is such a phenomenal success out of PSoC, and what we decided to do is it's become so important that it needs divisional focus. That's what I'm going to talk about.
So my quick agenda here. I'll talk about the explosive growth at TrueTouch that's driven for us [ph], and specifically about [ph] Gen 4. The growth has been based on Gens 1, 2 and 3, and Gen 4 is far and away the best solution frankly in the industry. And so we just introduced that. That is the clear leader going forward, and we expect it to drive further market share gain for us. So I'll spend some time talking about that. We also have a guest speaker from QUALCOMM here, a key partner for us in the handset space, and then a short video from NVIDIA, another key partner, to kind of emphasize how we're growing beyond just the devices we sell. And a new [ph] division is the optical navigation technology we call ONS and our trackpad products. Those are all very, very complementary to, of course touch and CapSense. They're essentially different versions of user interface, and I think it's important that Cypress is now kind of in a unique position of being able to deliver complete user interface solutions, not just touchscreen or CapSense but also optical navigation, trackpads and other opportunities going forward. So that's certainly one of the other reasons we're making the change organizationally.
One way to grow fast is to pick the right market. This is a quick example of a market growing very, very fast. If you look at just 2010 to 2011, the market success in touch solutions, driven primarily by mobile handsets, has more than doubled, and so that's driven the results and the phenomenal growth for the key players involved, certainly including Cypress. We expect to go over 1 billion devices a year within 3 years, and Cypress is very well-positioned to continue to grow with this growth here. In fact, if you look at the last year, the market grew over 2x, and as Brad indicated earlier, we had given guidance at the end of the year between $230 million and $250 million. We're very comfortable saying we're going to exceed the top of that guidance quite comfortably. And in fact, while the market grew a little over 2x, we're going to grow pretty close to 3x this year. So we gained share in 2011, very proud of that. We did it primarily frankly in the handset space, the low to mid end of the handset space where we've tended to dominate, and that's driven our success in the space of people transition either feature phones or smartphones over to a completely touch interface. That's been a huge success for us.
Going forward with Gen 4, we -- with a higher-performance solution to [ph] market, we see great opportunity to expand in higher-end markets, but in the higher-end market with Gen 4, we expect the momentum to continue. So we're very proud of this success. It's been great year for us and we're excited about the opportunity going forward.
So I wanted to talk to just a few slides on Gen 4. Gen 4 is not just a new chip. It really is a complete new platform, and we made the investments and took the risk to not just design the next-generation chip but to redesign the technology it was based on to integrate a brand-new processing platform, the ONS [ph] platform we talked about, and redo the entire analog base. And it not only proves what we are absolutely sure is the best solution in the marketplace for handsets, but it's absolutely the same platform that can bring the same benefits to tablets and to larger form factor solutions as well. We made a big introduction about this at the Embedded Systems Conference. We've been sampling since the summertime. We're actually ramping our first customers and we're ramping production in Q1. And so it's a real -- it's real and it's a significant step forward for the industry.
I do want to talk about some of the key differentiators in addition to going into an ARM platform. First and foremost, this may sound simple, but we've changed this product into the base technology to support 10-volt on chip. All of the solutions in this space run 2.7-volt drivers to the external marketplace. We can talk about all the fancy circuit technologies we do in the world, but one thing we know for sure is that the signal-to-noise ratio solution is directly proportional to the volt that you drive to the screen [ph]. By just going from 2.7 volts to 10 volts, we have a 4x advantage over anybody just from pure process technology change. Frankly, this is one of the differentiators that make me comfortable about our position long-term in this space. Very few people in the market can go invent their own technologies and apply them to markets like this. We can take things like 10 volts and give us a huge advantage to our customers just by doing the transistors differently ourselves. This is going to be hard for people to respond to frankly, and we've already seen that our customers are more than happy to see the noise go down just by simply programming the output voltage to run 10 volts. So it's a huge fundamental advantage right off the bat.
Now, to go with that, I don't want to just talk about it in terms of process technology. Because we've been in this industry about 7 years, these are very innovative circuit things that are equally important in terms of reducing noise. One thing I want to emphasize is when people talk about performance in touch, it's all about noise. It's all people really care about, in that noise turns into higher performance in terms of response rate, lower power when you have great signal-to-noise ratio, and it also turns into cheaper displays that would [ph] stack up, they can do things like in-cell, on-cell sensor online [ph], but limiting [ph] all those cases is how well you perform in the presence of noise. As the suppliers go toward solutions that are thinner, noise gets worse, and how you perform in noise is the whole issue. Now it's the whole focus of Gen 4.
So this other thing we did, we call Display Armor is actually a combination of 2 very unique methods. They're essentially like, one of them, you could think of them like your noise canceling headset. It actually listens to the screen and subtracts the noise out of the system and reduces the noise that's coming from the LCD and other things so that we can pick up your finger in a very noisy environment. The other thing we do is not just listen to noise but we actually detect and synchronize with the LCD. The LCD at regular periods drives the panel. We can actually figure out when that is and choose to listen when it doesn't drive the panel, another way to really innovatively reduce the noise in the system. This 10 volts and Display Armor really could enable technology you hear a lot about like in-cell, on-cell sensor online [ph]. This is, of course, like many of our stuff, all patent-pending technology.
And finally, one other point I want to make, sometimes some of our competitors I think just accidentally make this mistake, they may say, "Hey, you've got 10 volts, that's going to be higher power." Power is really important in this space. In fact, actually the exact opposite is true. The fact that we drive 10 volts actually allows us to receive a signal faster, and we've actually produced a device that is 50% lower power than anything in the world today. And the reason is that 10 volts -- in the actual device, most of the powers consumed in the receive portion of the device transmits at very low power into high impedance. We transmit at 10 volts. We can wire the firmware so that you receive very quickly because you can pick up the signal much easier from 10 volts than you could for 2.7 volts, and then the result is, in a real-world device, much lower power. In fact, the 2 milliwatt active power is a factor 10x than our previous generation. So it's a huge improvement; at the same time, we improved the other aspects of performance.
So we think of this Gen 4 as absolutely a revolutionary step forward for us. We've had tremendous success on the first 3 generations. We expect this to drive even further success. And of course, while we brought out the device, we talked mobile phones driving the biggest portion of the industry. The fact is, that same technology is equally important in other form factors, and if you think thin, low cost and noise or 10 volts is important in small screens, it's even more important in tablet form factors and larger screens. So the Gen 4 platform is going to position us to do very, very well beyond just our growing success in the mobile handset market.
So want to give you a quick summary on that. We're really excited about that, of course, but one thing I also want to say is it's not just about having the best device. You also have to partner with the key people who provide the whole solution. And so one thing we do want to take a few minutes to do. We have 2 speakers, one of them in person, one of them on video, and from QUALCOMM, Senior Product Manager Leon Farasati is here to talk a little bit about the partnership we had with them in reference to their recent Snapdragon mobile phone, if you would about that partnership.
Thank you, Norm. Thanks for having me here. So as Norm mentioned, my name is Leon Farasati and I'm the Product Manager in the QCT business of QUALCOMM, so the chipset division where we have the Snapdragon chipset. Quick about what I do, so I manage all our Snapdragon mobile device platforms. So these are the platforms or the devices fully integrated, latest and greatest features that we make available to the community for basically enabling application developers, guys like universities, research organizations, OEMs and ODMs to develop new applications, new user experiences along our Snapdragon chipset. It's a program that we initiated about a year or so ago. It's been very well-received. It's been expanding ever since, and we're very excited about the whole program as a whole. And the feedback from the community has also been very great.
So why am I here? So at QUALCOMM, we engaged with Cypress to incorporate the TrueTouch Controller, touchscreen controller into these platforms. And from an opinion of myself, as well as everybody else in our product engineering division, it's been a great partnership, both in terms of the technology itself and also the support that we've received from Cypress. And so that's why I'm here to basically give the word in and I really appreciate the help and support that we've been getting from Cypress.
Now in terms of technology, there's a lot of advantages to the TrueTouch Controller that we've used. I picked 3 examples of some of the things that we've been able to put on the market. One of them is, and I think Norm mentioned is, in terms of noise immunity and performance of the controller and the fact that it doesn't impact performance. So we have all these great features, fast end times and noise immunity, and this is really important for the developing platform, because the platforms that we're sending out, these have next-generation chipsets, and the intention is for us to showcase the capabilities of the chipset. So it's really important for the touchscreen to be responsive and robust so that we can really bring out that performance and the feature set of the chipset itself. So that's been very important.
In terms of features, we've also -- another area that we've been able to really take advantage of is the ability to have multiple access touches on the screen itself, and that's really important in all these new use cases: a lot of people doing gaming on their mobile devices; there's also things like electronic pianos where you want a lot of touches and you want multiple fingers on the screen. So that's been an area that everybody is excited about, and there's a lot of applications and use cases obviously that the developers are building up. And then the last one I mentioned is the idea and the technology in the TrueTouch on palm and grip rejection. So -- and that's also very important especially as you go on devices like tablets where you're holding your device and you have your palm on there and you're gripping it and you don't want that to interfere with the user experience. So that's another area that I think has been great.
So in closing, like I said -- I didn't bring any slides, I just wanted to talk a little bit about our experience. And it's been great, and we're really excited about the Gen 4 and incorporating that into these devices in the next generations that are coming out, and again, very excited about the technology and also very happy with the support that we've received from Cypress. And we're looking forward to extending it, and building on top of that. So thank you very much.
Norman P. Taffe
Appreciate that. And I just want to show a real quick video from Bill Henry, who's the Director of Product Management at NVIDIA. He just announced the reference design of the new Tegra platform and he's going to discuss that right here.
Norman P. Taffe
I want to point out, the Tegra reference design too is actually at the back table. There's a very cool game to it [ph], and that actually uses our TMA 84, which, as kind of he pointed out was the first single-chip solution for tablets, and that puts in that platform, which is a previous generation but the first single-chip large touch platform, so take a look at that.
All right. I'll finish up with a few kind of summary as we look at some of the markets. While we've been talking primarily about the mobile handset space, we've had -- we certainly are growing our large screen business significantly. If I look at that space outside of the Apple tablet space, obviously, there's been a lot of uncertainty, let's put it, where people, like, build up a lot and it's not sure if they ship through very many. There's some opportunity there for business but it's not real long-term opportunity. We do see that changing. We do see that opportunity becoming more stable over time and becoming a significant part of the market. We're really committed to single-chip solutions. We all love the [indiscernible] chips in the system, but at the end of the day, this market doesn't support that very long, and so what our strategy is that as we're the first to market with single-chip solutions, we did that with the TMA 84. We will be soon announcing more single-chip solutions for larger and larger form factors that are based on the Gen 4 platform I talked about. And as I said, the signal-to-noise advantages, Charger Armor, Display Armor, 10-volt drive are just as important, if not more important to the tablet space as they are to the handset space, so this is going to be a significant growth factor for us going forward.
Often lost in the mix, what you folks hear about touch is this is a very significant base business which continues to grow for us and for the user interface, not just CapSense buttons [ph]. It's a space where we're far and away the market leader. We were not the first there, but once we got there we built the best solutions, we were able to grow with the market, and we are now 3x larger than the next biggest supplier in CapSense buttons. So that's something that we kind of forget about, but we continue to find more and more [indiscernible] as this continues to grow for us. A great base. It is the base of the technology that turned into TouchSense, and we continue to innovate. We introduced something last year called SmartSense that actually senses the capacity of the panel itself. You don't have to tune it, it tunes for you, and that's the way we've been able to expand the market with CapSense dramatically.
You saw this earlier, albeit you may have seen it with different numbers, this is the IP position. I put it in here as well, as Dinesh did, but I re-sorted the table differently to make a different point. We had a significant patent portfolio in the PSoC space, but if you just divide that portfolio up, you'll find that now we have over 200 patents in the capacitive sensing and touch space, either filed or granted patents. So this is a big issue with customers, particularly recently. They want to make sure that if they're using your solution, are you protected? Have you invented your own IP? As lots of people jump into this space, usually the answer is no. But Cypress's answer is actually yes, and so from both the customer support side, it's very important for us to be able to say we have this capability and it's getting selected; at the same time, we are not afraid to use it either. So this is a patent position that we expect to use to our advantage going forward and make sure that people are going to jump in, they'd better be inventing their own stuff or else they'll hear from us. So I want to take a little bit of different, I guess a more aggressive stance, maybe Dinesh did.
A couple of things I did want to talk about is there's 2 new elements in the group. One of them is the programmable optical navigation we've talked about before. This is now starting to take significant traction for the company and growth for us. You'll see in the back handsets from RIM and Samsung which have our optical navigation capability in them, and so it's a growing opportunity for us. It is programmable. It's a very innovative solution. It allows for navigation with the finger and very complementary to CapSense and TrueTouch. There's a huge amount of phones that still use mechanical navigation capabilities, particularly in the middle end to the low end of the phone space. We think that Optical Nav is a better solution for all those, and with the ONS group being able to integrate to the other capabilities in the phone, I think this is a great opportunity for further TAM expansion for the division.
One of the things that's very important is our solution is quite versatile, so today, some of the times when we produce it just does the Optical Nav function. In other applications, we have capacitive buttons and Optical Nav, and we even have designs where we can do a simplified version of touchscreens with single touch solution where we do the capacitive buttons, a small touchscreen and the Optical Nav. We are unique in being able to bring those different technologies to market together. That's one of the other reasons for integrating them in a single division.
And then the last and newest piece to the group is the trackpad business. This is a level [ph] business that was really run as a startup as we're getting into the trackpad side of the market, but it's another significant TAM expansion opportunity. If you think of trackpad, it's really nothing more than a touchscreen on copper, on a copper floorboard. It's more into the different market space in terms of the PC customers, primarily, but now it's tablet. Those are the same customers that are also looking at touch solutions. And what customers are really looking for in this space is touchscreen-like performance. We've been working on this for a couple of years, and we've now penetrated some major accounts. Recent Dell introductions, Lenovo introductions are used on our touch -- trackpad, and in the coming years, we're going to see substantial growth in this portion of the business as well.
When we differentiate this, how to win a market that's, in this case, has been around for about 20 years, well, allow me to reiterate. We get into the market when there's a change in the market, and the change is really the trackpad going from a single, what we call TMG or multi-touch gesture market to multi-touch all-point gestures, just like you see on your touchscreen. So we were able to take our innovative all-point technologies and apply them to trackpads, and there's a dislocation [ph] in the market. The old trackpad suppliers couldn't all do that. Some could of course, but we're able to bring that capability into this space, and primarily, most of our design wins are people wanting to adopt that higher-end capability like you see on our new Apple trackpad. So that's what we really utilize as the way to get into this market and it's a very, very big damn [ph] opportunity for us.
So in summary, the division that we have now is really -- has a very comprehensive user interface portfolio. We have leading position on CapSense and touch products. To that we've added the optical navigation and the trackpad business, and because of the success we've had, we've really had the ability to fund further investment in user interface in general, both in terms of protecting our own IP position but also in terms of finding new solutions and combining the technologies to be differentiated. And then secondly, most important point is Gen 4 is here and it's going to drive further market share gains. We have a clear leadership and performance not just in the handset space but really in all touch, and as part of the investment we put here, Gen 5 is well underway. We now have kind of parallel development efforts. Gen 4 is out today. We're going to keep raising the bar. It is not a market that's going to be easy to get into because we're going to make sure that we keep making it harder and harder for the competition.
With that, I think we're on to questions. Brad?
Brad W. Buss
Thanks a lot, Norm, I appreciate it, and we're just a tad off schedule and it almost feels like an earnings call. I know you guys would like that. All right, so we're going to open up generic Q&A to anybody, any questions?
Unknown Analyst -
Norm or TJ, what was the actual changes, significant changes in the redesign of Gen 4?
Norman P. Taffe
In terms of [indiscernible] it's kind of across the board, first and foremost, the technology itself, we use -- our channel technology [indiscernible]. We added 10-volt transistors to it to drive the 10-volt signals, and normally what [indiscernible] is like, that does not require a customer to drive 10 volts to emphasize [ph]. We did internal circuits, they provide their normal 2.7 supply, we pump the voltage and we drive the signal forward, and that's all done with no change to the way they do a handset. So that was fundamental number one, and therefore the 10 volt drive should drive that. Number two, as we move to the ARM Cortex and zero core. We had 32-bit performance in the lowest-power CPU core you can get. So that was the second, obviously very significant change. And then third, where actually most of the engineering went into is a complete new analog channel. The analog channel included what we call listening channel, which is an extra channel where you just listen to the screen and then be able to detect and change your operation based on the one that you see on the screen. Those 3 things I guess the most fundamental, but it was a complete platform change, which I think puts us in a position to grow off that platform and make derivatives factor than we did when we made decisions on Gen 4.
Unknown Analyst -
My question is going to be for Dana. TJ might opine on one of them. Dana, you put up [ph] hoping to reach mid-40s for market share SRAM. I'm wondering where do you go after you reach that given that was the cap or ceiling of market cap that the industry put on Samsung. Second to that, will Samsung exit the industry in whole or would they still be supporting their key customers by the end of 2012? And finally, could you help us understand in the lawsuit with GSI where the company intends to go forward with that in terms of if there is no license or any working with GSI and you just eliminate them, logic says you're not likely to get their full share on a whim [ph]. And just wondering what was the process and what were you going after and has there been any damage to customer relationships in the filing of the ITC case?
Okay, let me see if I can remember all the questions. The first was the market share. So after mid-40s, higher than mid-40s, so we have an internal goal to get over 50% but I believe we should be able to hit more than 60% market share. It's a consolidated base. There's not 20 different players, so I think we can go -- you said there was a cap. You mentioned there's some kind of cap.
Unknown Analyst -
Yes, well, there were more players in the market then, and we are -- they had very strong competition from us at the time. So we think we'll be able to distance ourselves probably more than they did, although they were quite formidable. The second question was, refresh my memory?
Unknown Analyst -
Oh, yes. So we don't have 100% intelligence on this, but it appears that they're staggering their EOL, but for the most part, they'll be out at the end of next year, and that's every indication we have. The default is they've taken last-time orders by midyear and last-time deliveries by the end of the year. And there's a little bit of a stagger between customers but essentially that's the norm. And then the last question with respect to GSI, maybe TJ wants to opine. It's not about market share, it's about protecting our IP rights, so you saw today from different guys, we believe in IP, we believe in innovation, and to us, we don't want that compromised.
Unknown Analyst -
Question back to Norm...
T. J. Rodgers
And I'd add one thing on the answers, with regard with the shareholders want is growth of revenue and profit, and the handset market has oscillated between $1 billion and $5 billion, now run at the $1 billion level since 2000. We had a discussion about DRAM versus SRAM today, the fact that DRAM bits are 10x cheaper than the SRAM bit, and anytime a DRAM bit can take an SRAM bit it does. So SRAM's kind of a base where you can articulate why no other chip than an SRAM can be used, that's $1 billion, and then about every 5 years there's something hot that goes on. In 1995, the cache memory in every personal computer was an external SRAM, boom, took it up to $1 billion. That went away and got integrated. If you look at an Intel chip today, there's more SRAM area than there is microprocessor area on the SRAM chip. Then we had cell phones and then that went away and then we have routers. That is unlikely to go away completely. Some of it has gone away. The packet buffer on the router, basically the memory that's a giant pipe [ph] that stores data coming in on one end and peels it off the other end as it routes it, that has gone or is going towards DRAM because you can use the DRAM because how it is used is completely deterministic and therefore you can hide the inability of the DRAM to random access efficiently. There are other parts of the router where that is not the case and SRAM will survive. So we see SRAM as being $1 billion business with an occasional spurt, which I have never yet once predicted when I've been in the SRAM business literally since 1975 in 2 different companies, 3 different companies including Cypress. There will be some room in the future, but Dana talked about a non-voluble SRAM, we didn't have time to talk about it. But what about an SRAM that's as good as any SRAM in the world and it has a little capacitor on the board. You get one more out of the 32-bit or there, and when the power goes off, the capacitor holds enough energy to keep the data in the SRAM and store it in non-voluble bits like flash, literally flash. So that market is currently what, $30-ish million, $50-ish million, but it's growing and it's going to get more important, and if we can bring down the cost and bring up the technology level of that technology, there's another place for us to go. So I view our SRAM division more as a collection of people with memory, yield and technology skills that can apply those skills to new places. So the SRAM is going to -- we're looking to take market share but that's not really the name of the game for that group. We will work on new things to grow that group's revenue and profit over time.
Unknown Analyst -
Norm, if you could, first of all, on the low-end touch business where you've been particularly successful, can you give us an idea of what you're doing both in terms of smartphones and tablets to protect that position, particularly e-readers, where you've done really well? A competitor recently announced a design win on a Samsung platform that we would've thought would have gone to you, and the other thing is if you could give us a sense of how we should see sizing up your opportunity in the optical FingerNav and the trackpad business.
Norman P. Taffe
Okay. On the first statement in terms of smartphone, I think we've dominated the lower bit of it. I don't want to understate -- we have a huge business in smartphone. [indiscernible] You don't get as big as we got without really covering the entire market. I think there's a certain portion of the market, like you said, that we have been particularly significant. As far as protecting that business from other competitors, the way we protect it is we keep delivering better solutions. Gen 4 is, you might say "Well, it's the high-end solution for mobile phones," but the fact is Gen 4 makes every phone cheaper because it reduces the cost of the sensors, allows them to go to less-expensive solutions, and they really take more money out than the whole cost of the chip. The fact that we have [indiscernible] is a huge success is at the very low end with something called low-cost sensor, a single layer technology and our sub cap has turned into a new market share gain. So while we're excited about Gen 4, our Gen 2 solution is actually going to grow significantly next year in touchscreens because of that, that it allows them to take off doubt. So I think what Gen 4 brings us is the very highest end as well as the solution all the way across, and we are going to compete in every part of the market. I think one of the strengths of Cypress is, once we get our hands and our claws in the market, we can compete with anybody and make money out of it. SRAM is a perfect example. CapSense is another example. USB is another example. Once we get in that position of IP, we have the fundamental strength in terms of cost and tenacity to protect that market top to bottom, whether it be [ph] IP. But more than usual, it's just good cost and structural capability to address any price in the market if necessary while still innovating and adding value so we can keep our margins high. So I think from that endpoint, there's nothing in the market that I feel we are going to be able to defend and grow into. The second piece was related to ONS and trackpad. I don't know if -- I'd like to talk to Brad about any -- I don't know if I can give any guidance per se. Both of them are at the very front end of their growth. So they're going to grow potentially on a percentage basis. Actually, in 2012 it affected a really meaningful impact to the overall number. They'll still be small relative to touch, but they'll be growing at a very high pace because we're in markets where there's not many competitors, and therefore it's completely a loose end for us. So they're going to start to be meaningful this coming year and we see them being much bigger after that.
T. J. Rodgers
Okay. First of all, at the back of the room is [indiscernible], raise your hand please. She is working on a project where she's hooking up an RF chip made by Cypress to a dongle made by Cypress to a fermentation system, 152 of which we've given to UC Davis. I sent her an email earlier to ask her to sneak into the auditorium so I could tell her that I wanted to have a meeting in the boardroom at 12:30 with my group that's working on this project. She decided to sneak into the back of the room, hopefully not to get noticed but it's turned out to work exactly contrary to her plan. So I'll turn you loose, 12:30, see you on the boardroom, bring your team. Next question. Up here.
Unknown Analyst -
So just a question on alternative technologies to capacitive touch such as optical, do you see that as a potential threat? There are smaller companies like Neonode that seems to be making some entrants into that market. Just maybe if you could comment on that. And then a question for Dinesh about the implications to some of the other microcontroller...
T. J. Rodgers
Let's do one at a time. You guys have been stacking up 4 and then you actually have to reask the question a second time. So Norm, grab your chair. Norm, grab your chair. First question is about other technologies other than touch. I will answer that question.
The reason touch technology is taking over is because it is sensitive and it allows for robust computer control of multiple interfaces, 10 fingers, palm rejection. By the way, the drooling baby problem, we now have a camera that's got a touch control on it that works underneath saltwater. Sony makes it, saltwater. So why is that taking over? Touchscreens have been around for a long time. They are resistive touchscreens, and basically when you push something, it connects 2 conductors, but you just can't get all the information out of it. Our touchscreen, an expensive one, really expensive one is...
Brad W. Buss
T. J. Rodgers
Brad W. Buss
T. J. Rodgers
$2, and the cheap one is...?
Brad W. Buss
T. J. Rodgers
$0.50. Okay. Now you tell me about how these guys are going to take 12 sonic transducers, their 12 lasers, zoom the lasers around, track your finger, interfering with a bunch of lasers, run it into the computer and beat us to $12. Don't think it's going to happen. I've been wrong before, but right now, I think we've got the right technology, and the fact is, there are other good companies in this field and we will challenge each other to an accelerated race, an Indy Intel type race for example. And I would -- let me say it this way: we never invest my money to back a startup that was trying to bring in new technology into a space that's very well-served with the very cheap, efficient technology which had a Moore's Law type future in front of it. In the back, I saw a guy. Let's do 1 question at a time. Somebody is waving. Okay, next question. Who's got a question? Yes.
Unknown Analyst -
In one of your slides, your investment or your initial investment in Deca, can you talk a little bit about what are your plans in the future as far as future investments as well as your current ownership structure within Deca if somehow results are making investment on their side as well.
T. J. Rodgers
We've got $35 million invested in Deca technology. That is approximately what's required to put together a line. They've burned some on the money on people. We intend to invest, according to the current view of the plan, some $85 million before they will start making up gross margin that become cash flow neutral and we won't have to invest any more. We are the majority investor, and therefore from a board governance voting point of view, the controlling investor, we and SunPower are the equity investors in DecaTech and DecaTech's employees. Our typical plans for startup, I don't know the exact number for DecaTech as we speak is to, like in the venture world, let me post that as 20% of the company, and therefore, if they can make the company worth $200 million, 20% of that is $40 million, knock yourself out and make as much money as you can, please. That's the way -- the plan for DecaTech. Next question. Right there.
Unknown Analyst -
Now that you've broken out some of your segments, you have SRAM already, can you talk about what percent of revenues the touch segment is, USB and feedback, and more importantly kind of what's the growth rate you expect for those segments?
Brad W. Buss
Unknown Analyst -
Not yet or no, Brad?
Brad W. Buss
No. Well, probably [indiscernible].
Unknown Analyst -
Going to the -- moving to a different division, I want to focus on the PsoC 3 and 5, it sounded like you guys maybe had some holes in distribution in terms of that Tier 2 you were talking about. I wonder if you could maybe quantify if you're going to add more distributors or any other changes to the go-to-market strategy, be it more salespeople to that market? Any color on that would be helpful.
T. J. Rodgers
We haven't had a hole in distribution. We have future err [ph] on Avnet, so we've got big ones. We've got a network of Asian distributors. What we've discovered is that, we knew this with our customers. The first PSoC design is a hardline, it's a paradigm-changing chip. It is a chip where you in effect have to design your own chip first and then use the chip that you've designed to solve your problem. It offers a huge improvement in flexibility and co-engineering of hardware and software, but there's a barrier to that first chip, after that first design rather. So what a surprise, we focused on the top 10 of the world and we gave them 3 or 4 of our engineers and helped them get over the hump. And there have been customers, shining examples of -- our nonstrategic customers, our smaller customers who've gotten over the bumps themselves. But what we really want for the second division is to have 20,000 -- part of our corporate core values, 20,000 PSoC customers that are all designing PSoC and are enthusiastic PSoC supporters, and you simply can't do that job and simultaneously feed the top 10 cellphone manufacturers in the world. So we divided it so we can put more effort, less on distributions -- they've been there, they've been ready, we haven't been helping them enough -- more on solutions. Here's the 5 kinds of battery chargers. Here's how you do wireless interconnect for this classic device. Here's a kit we call MFi, made for iPod, iPad, where you can create your own products. The one I love the most is the oscilloscope. It wasn't called that but it was an oscilloscope on an iPad. So where we're going to put more effort is put more of our engineers and brain power behind bringing out a huge array of solutions in many, many different markets that are canned and ready to fly, so you can make what you're going to make, a boombox that's powered by an iPod, for example, and get it out there and then get over the barrier and then become a PSoC supporter. So it's just a bandwidth issue and it's really on our side. Our distributors are doing a reasonable job. They are dedicating people to all of our major distributors. Matter of fact, almost all of our distributors, major and independents are dedicating people directly to PSoC. It's a problem of plenty. We've got way more opportunities than we can properly service, therefore I need 2 divisions, and as all companies mature, they stopped being organized around what they make and the internal focus is to start being organized around who they serve. And what we really know is we've got 2 PSoC divisions. The chips they use oftentimes are the exact same chip. They have different numbers and they serve 2 different end markets. So we're starting to become more customer-focused. We've been an engineer, internally customer-focused too long [ph], that's why we invited some of our customers here today to show you that we truly mean that we're going to change that.
Unknown Analyst -
So TJ, I guess judging by your earlier answers, you're not going to throw a target for PSoC revenue next year?
T. J. Rodgers
Not yet. One of my favorite things, as I used to be in the forecasting business and I just don't do that anymore and I leave it to him, and I'm happier and he's happier but thank you.
Unknown Analyst -
A different question then. You talked about in your opening monologue about you're not sure that we're at the bottom. You had a few semi companies out there, some distributors, even some [indiscernible] call for a bottom here. Maybe you could just take us through your business trends over the last 3 to 6 months and what signs are you looking for? What have you seen in previous cycles that "Okay, I've seen this, we're definitely at the bottom"?
T. J. Rodgers
First of all, we look at the general market, and we are not necessarily a proxy for the general market. I'll add that the Director, Dan McCranie, who's also on the board of ON Semiconductor, so what are my leading indicators is I say, "So how are things going, Dan?" Then he tells me, "We have 2 groups that do reflect the market at large: our asynchronous static RAMs." Asynchronous static RAMs are high-tech stuff for routers. Our asynchronous static RAMs executes memories for everything in the world that use chips. And we watched the sales of those. Right now it feels like we're at the bottom. It doesn't feel like things are pulling us down further, but on the other hand, other than our success with Gen 4 and in design wins, it hasn't shown off revenue yet, and the new USB 3, which is going to kind of spark all USB companies, including ours, I can't tell you a real rousing story how Q1, for example, is going to be a barnburner, Q4 will be -- we're on track for Q4, and Q1, now that lead time's shrinking, you're back to being invisible. It's like you're driving your car and your headlights shine out 250 feet. Well, if you can't stop within 250 feet and there's a brick wall at 350 feet, then you're going to smack into it and you just don't have any visibility right now. And I really do mean that we have not been able to forecast. Best news I've heard in a while, TSMC says they're going to have a better fourth quarter than the third quarter, meaning they're making more wafers for our competitors, meaning our competitors need more silicon and there's only one reason they need more silicon. I would put that factoid out more than anything I can tell you from looking at our own internal database.
Brad W. Buss
Yes, I think the big thing is lead times, right? I mean, our lead times -- Chris, why don't you give a little...
Christopher A. Seams
When we talked just a few weeks ago was, we had almost everything in the company below 6 weeks and I can tell you in the last 4 weeks, everything is below 6 weeks today. If you took an average number for the company, it's 4 weeks, so my brick wall is four weeks away. Last time I got on my calendar, which was this morning, 4 weeks gets me just into December, so I think it's too early to call what first quarter is going to look like right now.
Brad W. Buss
Yes, I think nothing -- you're not clipping off anything, but I don't think quite frankly what the macro volume is right now, I think it would be hard for anybody. We get a little clarity in Europe. I think it changes things. I think what we hear from customers and distributors is there is demand. People are just cautious and are kind of waiting, right? And hopefully we all get clarity. I mean, the governments are screwing all of us around the world. I'm not trying to sound like TJ or anything. I'm getting ready to move back to Canada pretty soon.
Christopher A. Seams
The question was, had we seen that stabilization in bookings that some of our competitors have talked about, and in every product line where our lead times were all below, the answer is yes. Where our lead times have continued to drop, the bookings rate are matching that drop in lead time, so I expect that'll flatten out too.
Brad W. Buss
There's a beauty going into the cycle. I love the fact that lead times are short, it just means inventory, they've done a good job of managing that down. We don't see pork belly inventory with any of our real big customers or end markets, so that's a good thing.
T. J. Rodgers
The one thing we do know is we're very careful about -- first of all, we only recognize distribution and inventory when they ship it, so there's no such thing as stuffing the channel with regard to our revenue but we're very, very careful about even over-inventory and [ph] distribution. We're now watching our shipments and we're actually working with corrective actions if we see the inventories go out of line in either direction. So right now, the one thing I can say about Cypress is that we're keeping the inventory chain and inventories pipe front to back, so we're not going to get surprised and say, "Oh, by the way, the good news is they're buying. The bad news is they're sending out enough stuff that they won't need anything for the next 17 weeks but we'll see in the third quarter." That one will not happen with us.
Unknown Analyst -
It was great to see that you guys are finally putting some stuff out about the trackpad and giving us a little information and Norm talked about the relative growth we're seeing possibly from that division in 2012. Could you help us now with your go-to-market strategy there? You got a competitor who claims 65% to 70% of market there. Are you going for a land grab in that business and less worried about margins or are you trying to beat on technology?
T. J. Rodgers
I love some of the reactions. They feed me with the inflammatory words. If I only repeat them, it will only cause a reaction among competitors. Let me just tell you the truth about that. We weren't looking at it. We have a Manufacturing Vice President who's not here today who said "Gee, trackpad. That's -- just what Norm said, he said he's exactly right. It's touchpad on copper. If you look at it, it's an array of diamonds typically and it's exactly like the IPO pads on a touchscreen. He said, "Gee, all of the laptops got one of them and we ought to make one." So the manufacturing guy hired a couple of engineers in Shanghai and they worked on it, and it kind of didn't work and they worked some more and kind of didn't work, and then they learned about drivers and they kind of got that going. They got a couple of customers who said they wanted an alternative supplier. There's a lot of concentration in the markets, so we worked with those customers to make it work. And one day I'm sitting on my one-on-one and I'm going to beat him up on yield or costs or whatever, and he says, "We got the trackpad that works and Dell's going to buy it" and I go, "Okay. Slow down and tell me that part again." So all of a sudden, I got this little freebie. This is not one of Cypress's grand plans -- fund the company, see the vision, give the employees equity, yada yada. Just how many people are in that group, Norm?
Norman P. Taffe
T. J. Rodgers
18 and that's way up. Trust me, that's 100% growth over where it's been. So we have a product. It's a very close cousin of our other products. The only additional thing we had to learn was the driver to the interface of the personal computer, and it works and it uses the same silicon other stuff does, and we've got customers that are interested. It's not because we're world leaders, we've got a product that's as good as anything out there, we believe, and they want an alternative vendor and we're glad to serve that purpose for a while. And it's a huge market, and therefore, we'll do our Cypress thing: We'll get to market. We'll ship it. We'll learn. We'll make it better. We'll ship some more. We'll try to get some single digit and double-digit penetration and see where life takes us. So from my perspective, it's a blooper, but now it's got our attention and we're going to get serious.
Brad W. Buss
Yes. I mean, we're not looking to -- we sell on value and we're going into the higher end that's evolving and changing. We don't want to land grab the low-end stuff that -- the form factors, as you all know, are changing, right? So they're either going to have some level of high-end user interfaces that are going to die, and we're going to be there if that migration continues, and it's not a margin thing. We don't have any really nasty margin businesses anymore. We're not looking to add any.
T. J. Rodgers
Anymore? Many of you guys are trying to get triple triples and I am willing to grab a second question. Do you have it? No? Thank you very much for coming. Lunch is served.
Brad W. Buss
We have lunch. Demos will continue, and thanks again very much.