Advanced Micro Devices, Inc. (NASDAQ:AMD) Credit Suisse 19th Annual Technology, Media & Telecom Conference December 2, 2015 4:00 PM ET
Executives
Lisa Su - President and CEO
Analysts
John Pitzer - Credit Suisse
John Pitzer
Perfect. Clock is going backwards, so this means it's time for me to start. I would like to welcome everyone to this afternoon presentation.
It's my pleasure this afternoon to introduce Lisa Su, who as you all know is the Chief Executive Officer of Advanced Micro Devices. We have a fireside chat format for the next 20 to 25 minutes. If you have a question, please raise your hand. We'll get you a microphone, but I'll kick things off.
Lisa, you just recently celebrated your one-year anniversary being appointed CEO of AMD. I think it was end of October, beginning of November. It's tough to say that you've repositioned the company fully, but we're kind of interested in understanding what you thought were the major challenges when you took the job as CEO and kind of how you -- what you've done to address some of those challenges.
Lisa Su
Yes great. So John I am happy to be here. Thanks for having me and I think it has been a very interesting year.
I would say at our core, AMD is a technology company and a product company and so it really has been the last year and really several years have been around really focusing the company on what we're good at, developing great products.
When you think about how long it takes to really transition our roadmap, I think I feel good about the progress we’ve made. Certainly there were a set of challenges, the PC market was a little bit volatile in 2015, but I think we've done quite a bit to position ourselves well.
With the product portfolio, I am very excited about what we're doing in discrete graphics and the opportunity to reposition ourselves in the PC business and then in the medium term with our datacenter play. So those are the key areas that I've been focused on.
John Pitzer
Lisa, maybe we can dive a little bit into more detail because at the Analyst Day in May, you really kind of outlined I think three kind of key growth drivers for the company. One was gaming. The other one was immersive platforms and then the third one was datacenter.
Maybe we can take each in sequential order. On the gaming front, what do you think are the largest drivers for AMD?
Lisa Su
Yes so if you think about gaming, whether you're talking about PC gaming or game consoles or even embedded gaming, I think what's interesting is that it's driving a lot of computing and so whether it's to drive 4K displays or drive new virtual reality type headsets, there is a lot of computing usage both on the CPU and the graphic side and I think we're unique in our ability to combine those two assets.
So we've been very successful with game consoles. I think the game console cycle with Sony and Microsoft, this cycle has been very positive and we believe that that extends definitely into PC gaming as a growth market and then going forward into broader gaming environment.
John Pitzer
Lisa, I know you get asked this question a lot in the gaming console side, but how should we as investors think about sort of the third year of a console cycle and whether or not you can continue to show good growth.
You've clearly kind of extremely well on the market share front and you had a good ramp off of the initial launches, but how do we think about sustainable growth from here?
Lisa Su
Yes, so I think the way to think about gaming is the game console cycle is usually anywhere from five to seven years or usually in the -- on the longer side and this cycle has been really good. I think it’s been really good both, Sony and Microsoft have really marketed well through the value proposition.
The peak of game consoles are probably anywhere from the third to fourth year. We believe this year game console units are up relative to last year. From our visibility next year will be up relative to this year and it’s really about continuing to satisfy that amidst is higher for more gaming. So, I view it as a continued good market for AMD.
John Pitzer
Could you talk a little bit about your traction in the discreet market? We had NVIDIA earlier today on a fireside chat, kind of curious is how you kind of view your market share position today and as you bring fury to market, kind of where you think your share can go over time?
Lisa Su
Yes, so discreet graphics it's always been a back and forth between us and NVIDIA. I think it’s a market where it really is driven by the technology and the inflection points and so very key inflection point for us is if you think about it, discreet graphics have been on 28 nanometers since like 2012.
So, it’s been a long time in the same process technology. Next year is an important inflection point. I think moving discreet graphics to FinFET devices. We'll get a incrementally significant power performance improvement. So I think that’s one growth driver for us. Direct X12 and new games coming out in that area another driver.
So I think we view that discreet graphics is an opportunity for us to grow share as well as just content in the market is going up as well. So from my perspective, I think Fury was a strong launch. It’s the first graphics card. The first really chip in the market using high bandwidth memory. I think it's at the high end.
We expect to refresh our entire portfolio in 2016. So I do view discreet graphics over the next number of quarters as an important data point for what we can do in the market.
John Pitzer
Every year we start off the conference with a sentiment survey asking different questions of investors and one question we’ve always asked is what's the most investible buzzword for the coming 12 months and this year VR came in number two and I think it was the first time that virtual reality actually hit the screen, so kind of curious…
Lisa Su
What was number one?
John Pitzer
Number one was still AWS.
Lisa Su
Okay.
John Pitzer
Because you know Amazon taking over the world.
Lisa Su
Yes.
John Pitzer
I’m wondering if you can help us better understand your positioning in VR and then help me understand kind of the compute power that you need to actually have an immersive VR environment and what that could mean for ASPs?
Lisa Su
Yes so I think VR is a very interesting market and the way to think about it is that we’re at the very, very beginning of the market. So think about it as early days where enthusiasts are really trying to live the experience and the great thing about VR is it's almost like insatiable graphics demand because you always want to be a little bit more realistic, little bit more lifelike.
And so from a discreet graphics standpoint I think it’s something that we'll develop over the next five years. It’s not a one year cycle. It's not a two year cycle.
I think there will be some progress in 2016, but we think of it as a long term growth driver for computing in general by the way both CPUs as well as GPUs and the nice thing about it is it does a cause a mix to the high end of content.
So if you look at Oculus, one of the leading VR headset manufacturers, they publish their minimum specs and these are pretty hefty GPUs that are needed. Let's call it $300 and above and it seems like people are actually trying to future proof their purchases. So I think all that's good.
John Pitzer
And relative to the $300 and above price plan, what's the average selling price within your GPU product portfolio today?
Lisa Su
Well I think depending on where you are, if you are in the notebook side, you're on the lower end and in the desktop side higher, but certainly the $300 and above price point is considered enthusiast class graphics.
John Pitzer
And then Lisa when I think of VR, I think it's more gaming than immersive platform, but relative to the three buckets of growth you talked about at the Analyst Day, where does that one fit and maybe we can transition into the immersive platforms and get a better understanding of what you really are trying to define when you say that term?
Lisa Su
Yeah so there is a bit of mix of the two, but I think we should look at it as I think gaming is the first sort of usage of virtual reality platforms, but if you put virtual reality and augmented reality together and I think the idea is that there should be a number of other areas where you want more lifelike experience than you have today and all those require moving pixels around, so driving graphics and so on and so forth.
So I think on the immersive standpoint, whether you're talking about VR or AR, it's looking at how you might improve productivity in industrial applications or other areas.
John Pitzer
And then on the datacenter front one of the key words before we had Xilinx on stage earlier, Intel earlier as well, but acceleration as a new sort of workload/application of the datacenter. As you think about attacking the datacenter space, what workloads or what parts of the market do you think your IP is most well suited to go after.
Lisa Su
Yeah, I think for the datacenter, there are really two things that are necessary. I think the first is having a really competitive just basic core technology. So in terms of CPU capability, whether you're talking about x86 or ARM, I think there is a general good enough level that you need of a CPU performance.
So I think that's one ingredient, but when you add on top of that acceleration, I think it gives you an opportunity to really fragment the datacenter into segmented workload. So from our perspective, we do believe there is ability for us to really influence the cloud market and what we can do with the hyper scale guys.
I think being able to customize and customize both on the hardware side as well as on the configurations, I think we can do -- and the key for us has been investing in new CPU technology to get us competitive at the high end.
So it's a big investment, it's an important investment. It's one that we believe will be multi generational. So right now certainly Intel has dominated the market share by far, but if you think about the companies that really have the capability to put out a high performance processor, we feel really good about our capability to do that.
John Pitzer
But one thing that makes AMD sort of unique asset as you sit both on the x86 camp and in the ARM camp and I am kind of curious if you can give an update as to how that heterogeneous environment is playing with inside the datacenter especially with some of the hyper scale players there.
What are the advantages that they might see in heterogeneous architectures?
Lisa Su
Yeah, and I think I would broaden John heterogeneous to be not just around x86 and ARM, but around CPUs and GPUs and FPGAs and all of the above. So I think the interesting dynamic now is people are willing to experiment, particularly if you have the ability to control your own software stack.
And so from that standpoint, I think there are lot of opportunities. For us we're leading with x86. I think the x86 is a predominant market. It's the fastest time to money. It's what we're good at.
I think ARM has a clear place in the datacenter. It's probably a little bit slower ramp than most people have -- maybe some people have predicted, but certainly there is an opportunity there.
I think what customers want is clearly the ability to have optimized for their workload and so a combination of CPUs GPUs and accelerators will be part of that solution.
John Pitzer
When you think about your datacenter efforts, how important is the mixture?
Lisa Su
I think Zen is the first of a multiyear strategy. So again you ask me what are my thoughts around the company. I think AMD at our core, we're a high performance computing company and so Zen is a form scratch architectural design.
For those of us to do those it takes a lot of work. So it's a multiyear effort, but I think it's a multiyear effort that we can see coming to fruition and so what datacenter customers want from us is one, we want you to be competitive and two, we want a long term roadmap.
And so we've really talked about Zen Plus Zen follow-ons as a three to five year view of what's needed to be successful with datacenter.
John Pitzer
Can you talk a little bit about your Morre’s Law roadmap, talking about sort of that future proofing for your customers? I mean the digital market is a market where you need scale because the investment levels continue to go up at a non-linear rate. How are you trying to balance sort of your balance sheets and cash flow needs with also your needs to invest especially as it pertains to Morre’s Law?
Lisa Su
Yeah, no, I think it's a great point. I think the fact is that Morre’s Law is very much a key part of being competitive in the high performance space. I will say the dynamics are changing a bit and the dynamics changing a bit meaning and I think most people will say Morre’s Law is slowing down a little bit.
I think if you look at the gap between let’s call it foundry and leading edge technology, I think that gap is shrinking. I think that plays in our favor. I think for us what we’ve done on the investment side is we've very much focused on the places that we know we have secret sauce.
So CPU cores, graphics, that IP is the core to our business. When you look at some of the other areas, you can really leverage third party technology and we have really kind of rethought about how we do R&D in the company.
John Pitzer
One of the things that's been particularly challenging in your first year is that the PC market didn’t cooperate. It’s been a difficult year for everyone in the PC market. I’m kind of wondering if you can give us your thoughts as to where we are now.
Do you finally feel like we’re stabilizing at the market level and additionally how do you think your market share is playing out within the businesses?
Lisa Su
Yeah, so the PC market certainly has had its ups and downs and perhaps a bit more down. I would say 2015 was a bit challenging. There were a few with macro and what’s going on in China and perhaps some other Windows 10 transitions it was a little bit challenging.
I do actually feel though that it is somewhat stabilizing and the way I would put it is I think the PC market is a good market. Its 300 million units, how many markets do you have 300 million units that can use all of these processors?
For us, 2015 has been a bit of a clean-up year in terms of where we were. We were a bit out of position with our channel inventories and so we've gone through that. I do feel that that is for the most part behind us and certainly as we exit the year, we'll be behind us.
And then when we go into with 2016 is perhaps a market sentiment that says, yeah maybe it will be down, but modestly down. We’re coming out of a good product cycle. We have new products with Carrizo that we think is very comparative, compared to Skylake.
We have commercial design wins that we think changed the mix of our portfolio. So I do look -- I do believe that the PC market has for us seen its trough and we'll continue to build from there.
John Pitzer
Lisa as you think out of the next three to five years and you try to kind of plan the investments in the business, what do you think your footprint or market share needs to be in PC such that that business isn’t really constraining you from investing in other growth markets? What’s the target here longer term?
Lisa Su
Yes so if you think about PCs for us, I think we’ve been anywhere from mid teens, high teens, little bit different between notebook and desktop. The important thing is the PC market actually helps develop a lot of the core IP that goes into datacenter that goes into our semicustom markets as well.
And so from my standpoint we certainly do believe we can grow share in PCs, but more importantly I think we can -- we built a very strong leverage model across the company for the R&D efforts.
John Pitzer
Can you walk through some scenarios that get us to a path back to profitability or breakeven for the company? What kind of growth do you think you need to see in the PC market and/or market share to get us there and how comfortable are you with the balance sheet as you make this transition from currently where you are back to a breakeven or profitable position?
Lisa Su
Yeah so that’s a great point. Look I think as we look across the next number of quarters, I think there are a couple of near term drivers and then a couple of medium term drivers.
In the near term, I think the catalyst in the PC market are around let’s call it a return to normal inventory levels for us in the client PCs and then market share growth in graphics.
From what we see from a design win cycle, we’re at the right stage of the design win cycle. We’re at the right stage of the product cycle and that’s the next two to three quarter type transition.
And then as we move into the longer term, we have additional semicustom design wins that are coming into place in the second half of '16 and then going into the data center play in '17.
So it is very much product cycle driven for us. I feel that we’re well positioned with what we need to do. I think the data would tell us we are at the cusp of that product cycle and relative to the balance sheet, I think we’ve done a pretty good job of managing through the balance sheet puts and takes.
One of the things that’s been pretty important to us is to ensure that obviously the long term debt maturities are out into 2019 and we were able to complete a pretty good deal for us in terms of joint venture of our backend assets with Nantong Fujitsu Microelectronics. So I think that gives us some balance sheet above as well.
John Pitzer
Lisa can we dig into a little bit more detail on some of the comments you made about the semi custom business, clearly you had great successes on the gaming console side, you talked about a couple new customers coming in the second half of next year and some stuff and Datacenter in '17, how should we think about your engagement?
Can you talk a little bit the engagement pipeline for your semi custom business and what gives you comfort that you can hit your target? I think its 50-50 is kind of the target that you ultimately have as far as semi custom as a percent of overall revenue.
Lisa Su
Yeah, so I think as we look at our balance between let’s call it PCs and non-PCs being at 50-50, I think we were there, we'll be there end of '15 certainly into 2016. We’ll obviously like to see growth in both businesses. So, I will say that.
But relative to semi custom, I think the pipeline is solid and what gives me good visibility into the pipeline is people come to us to do things that they can’t do with somebody else.
So we’re rarely competing in semi custom against let's call it a standard ASIC play. People are coming to us because they want access to leading edge CPUs or leading edge graphics and that’s something that’s fairly unique in the industry.
I think as our portfolio becomes stronger with our graphics refresh and our Zen CPU core that will bring on additional opportunities in the semi custom space.
John Pitzer
Last quarter you announced the JV with the Nantong Fujitsu over in China as a way to kind of monetize some of your backend assets and lower some of your expenses.
I’m kind of curious Lisa, when you look at AMD, you’ve got such a wealth of IP over 10,000 patents, your very unique piece of real estate in the semiconductor market. What other avenues of monetization might you pursue other than building chips and selling them into the marketplace?
Lisa Su
Yes, so look, I think I take a fairly holistic view to what AMDs assets are. So our first and foremost priority is always around building and selling products and we continue to invest in the IP to do that.
But I’m also very cognizant to the fact that we’re not going to make every product that can utilize our IP. So I do believe that with our IP, whether you talk about patents or technology, there is an opportunity to leverage that both for additional partnerships as well as additional licensing opportunities.
I think that used to be something that we would do as a let’s call it if it came up and I view that as the next step of the evolution of our new monetization efforts.
John Pitzer
AMD is not a stranger to direct investment from third parties, [Shin] [ph] in China has been making a lot of o overtures of trying to get semiconductor IP seeing the [Western did Unisplendour] deal, would that be something that would be attractive or something you would be open to as far as ways to try to bring capital into the company?
Lisa Su
Yes look when -- when I think about our strategy, when our Board thinks about our strategy, we do believe that investing in our product and IP portfolio is the number one priority. So that is what we’re focused on.
As we look at obviously there is a lot of M&A going on, various pieces in the industry, I think the opportunity to partner is something that we'll always look at and as we look at how do we best leverage our assets, we’re looking to get our technology in broad enough areas. So that’s where I would view it.
John Pitzer
There are many questions from the audience.
Question-and-Answer Session
Q - Unidentified Analyst
Can you talk about what type of work you’re doing in the auto space?
Lisa Su
Yes so I think if you look at automotive, it’s not a space that we have selected from one of the embedded marketplace. So if you look across sort of the markets that use high performance CPUs and graphics, we believe in embedded there are Fin clients, Casino Gaming, a bunch of things. I think automotive is an interesting market. It’s just not a place that we have invested at this point.
John Pitzer
Lisa maybe I can -- Randy over there?
Unidentified Analyst
Yes just a question, you have the wafer supply agreement every year and the last few years it's fallen a little bit short of that. So I’m just wondering would that continue to be an ongoing commitment each year for a certain wafer supply agreement.
And then how do you see -- you talked about moving a lot of FinFET for graphics and processors. How do you see the mix between your foundries playing out over next year or two?
Lisa Su
Yes, so in terms of the wafer supply agreement that we have with Globalfoundries, it is something that we annually look at what our commitment purchases will be. So we are in the process of negotiating our WSA commitments and I think we'll continue to do that.
I will say overall that our Globalfoundries relationship is strong and continues to get stronger. So I think we're fairly aligned in what we're trying to do there and so we'll certainly continue to keep you updated on how that goes.
Relative to foundries and our foundry strategy, we have historically used both TSMC and Globalfoundries and we'll continue to do that. I think we've used -- we're in the process of using both the 16 nanometer as well as the 14 nanometer technology at TSMC and GF and we're actually very pleased with the results at both foundries.
John Pitzer
So I can ask a follow-up question on x86 versus ARM and that clearly I think expectations a few years ago that, that ARM would be -- have been a larger present to the datacenter than it does today. I think you've rightfully kind of set expectations at still a couple years out.
What's the impediment? Is it really the ecosystem building around ARM that needs to develop more fully and if that's the case, what can you do to help sort of invigorate that ecosystem?
Lisa Su
Yes, so I think it's a good question. I think there is a lot of interest in experimenting in the datacenter. So there have been a lot of effort on porting software to ARM. We have participated in that. Our Seattle ARM-based chip actually is shipping production in the fourth quarter in modest volumes.
And what you'll see is that the ARM-based chips out there are good for developing the ecosystem. They're not at the performance levels yet of x86 still and I think that limits sort of the market space that it can attack.
Now there is no fundamental reason ARM can't reach x86 performance. I think it's a matter of time and money to get there and so over a period of time, I think that will happen. I think ARM has done well in the networking and the storage markets, but x86 has such an install base and it's a strong install base and it's just -- the barrier to entry is just a lot higher with the new instructions there.
So again I continue to believe that x86 for us is a very strong play. ARM is a longer term play in the datacenter.
John Pitzer
Great. With that we've ended our time here in this session, but I want to thank Lisa and everyone for joining this afternoon. Lisa, thank you.
Lisa Su
Very good. Thanks John.
- Read more current AMD analysis and news
- View all earnings call transcripts