Seeking Alpha
We cover over 5K calls/quarter
Profile| Send Message|
( followers)  

Fusion-io, Inc. (NYSE:FIO)

February 27, 2013 4:55 pm ET

Executives

Lance L. Smith - Chief Operating Officer and Executive Vice President

Dennis P. Wolf - Chief Financial Officer, Executive Vice President and Principal Accounting Officer

Analysts

Kathryn L. Huberty - Morgan Stanley, Research Division

Kathryn L. Huberty - Morgan Stanley, Research Division

Good afternoon, and welcome to the Fusion-io session. I'm very pleased to welcome Lance Smith, COO, and Dennis Wolf, CFO, to the stage. We're going to jump right in to the discussion, and I'll make sure to leave a few minutes at the end and take some of your questions as well.

Question-and-Answer Session

Kathryn L. Huberty - Morgan Stanley, Research Division

So to kick it off, Lance, I want talk a little bit about Fusion-io's competitive advantage. There's a view in the market that with a lot of more component-oriented companies competing with Fusion, at least making noise in the market, that perhaps this is a product that just gets commoditized and maybe eventually embedded on the motherboard and the ability to make margins and differentiate goes away. So maybe talk about Fusion-io's competitive differentiation versus others in the market.

Lance L. Smith

Absolutely. So first and foremost, so we're a solutions, we're more than just a component. And what we've shown over time, especially when we first introduced the technology to the marketplace, that we've created a hardware platform, creating our own controller that actually interfaces with demand and then layered on top of that, software components. The first level of software components is our VSL technology that does the translation from NAND and representing the storage itself as a usable form for the operating system and the file systems and the end-user application. We continue adding on to the value proposition in that software stack. So we go from DSL, we add caching solutions. So we have direct cache that can be deployed on stand-alone service solutions. And then, of course, we have our ioTurbine acquisition that allows us to address caching in and applying flash on the service side for the virtualized server environment. Now we didn't stop there. We just recently introduced the ION software that actually allows a bar, really an integrator to choose an OEM class server, an HP, an IBM, Dell server and actually add our drives and add ION software to turn it into a shareable device. So you can actually have it sit in front of a SAN or a NAS or you can actually have it stand-alone as a storage solution. Now we've added one more component to that, and that's our Software Development Kit. But then this, again, helps us to get stickiness in the marketplace, helps customize solutions for applications and customers. And then software development kit is actually a method for ISVs, integrated software vendors, to customize their software, natively connect and address our devices, I guess, in the most possible performance. Now the SDK has all kinds of components, accelerators for very specific applications, file systems that we created very specific for flash-based storage solutions, and this help differentiate us from just a component.

Kathryn L. Huberty - Morgan Stanley, Research Division

Sure. So one of the concerns is that the controller specifically becomes commoditized. Does Fusion-io solution need to evolve if that plays out? And what does that mean for the Fusion story?

Lance L. Smith

Well, I'll tell you. Since day one, there's always been competition. There's always been some level of just good enoughs in the marketplace that started out first being 15,000 RPM hard drives and then it was SSDs. Then when you look at the potency of flash, it can come in a variety of forms. There's a need for large capacity and performance. There's a need for cost optimization. And then maybe you can actually take flash in its smallest form, maybe you can put it down on a motherboard. But what's common amongst all of them is that you need a technology to interface to it. Flash memory or NAND memory doesn't look like anything that our operating system recognizes, that any existing drivers recognize. You need that translation layer, and that translation layer is our VSL, our Virtual Storage Layer technology. Whether it's small amounts on a motherboard or large amounts on our ioDrive, you still need that IP to actually do that in high performance, high reliability, so you can go into mission-critical applications.

Kathryn L. Huberty - Morgan Stanley, Research Division

Okay. And one of Fusion's largest customer base is the hyperscale company and 2 large companies, in particular. And there's a view that those companies only use the product as a cache and, therefore, they don't necessarily need all the incremental software. Why did those companies use Fusion-io versus another good-enough solution?

Lance L. Smith

Maybe the use of cash is probably overstated. When you look inside a data center, temporarily using data where the permanent version of that data is stored on some kind of back-end store, there is usage models that place information local in the server itself, and I would describe these hyperscales in using 2 different models. They'll use a scale-up model, where they'll use big ION large storage amounts of capacity on a very large server platform, like we've seen customers who's taking these 4 new servers and they put tens of terabytes of flash storage on that single server to get the most possible performance out of it. And by the way, they put that much flash into single server because they want the entire database to fit into it to get the most possible performance out of the CPUs. Right? Because what we're trying to do here is open up the bottlenecks so those CPUs that were idle are now being fully utilized. Then there's scale out, scale out to the sense where they're trying to make use of the number of applications, the number of users that they're trying to support. So they'll put out lots and lots of servers. And all you have to do is put a Fusion-io product in each one of those servers, then it gives you this scale out or this scalable solution where the more servers you add with Fusion-io embedded into it, the more performance you get almost incrementally. So these are cluster-based, Hadoop-based. These are large scale-out deployments where you're trying to serve millions of users with the best possible service and reaction time of that service that you're offering to the customer base.

Kathryn L. Huberty - Morgan Stanley, Research Division

So just carrying on that point, there are some other companies that have certain earnings calls and they say, "We won big hyperscale companies, and we won those against Fusion-io." Why wouldn't a hyperscale company use a competitive solution? Why are they using Fusion versus the others? What the loss that they have?

Lance L. Smith

Probably the easiest way to describe that, when we start talking about performance and price performance and their value, there's 3 aspects to a storage that one has to be concerned about: bandwidth; IOPS, right, the read and write operations that the product can support; and then latency. It is that last attribute that's really the most important. We've shown that, like hard drives that support bandwidth quite well, streaming data, streaming video. Right? You can use hard drives to get that done. But then IOPS, there's this big IOPS race. And what was interesting about that was that people were trying to compete on IOPS and we always had high IOPS, but we needed to show people why the application itself was really scaling. When you put a product into it, you get 10x more application performance. Well, in IOPS, a year ago, we showed you can do 1 billion IOPS. I mean, it was no joke. We literally built the design, had the hardware, showed the demo, here's the 1 billion IOPS. That's not where the race is at. The race is at how fast can you get the data that the application is asking for. So what we do know, that if the application is sitting around waiting for data, you're not getting your performance. But if you can deliver it in milliseconds, very quickly, in nanoseconds, then that application will scale in performance and give you 10x or 20x or 30x improvement in your application performance. And that's where it's a challenge for our competition: to do all 3 of them, not just 1 of them.

Kathryn L. Huberty - Morgan Stanley, Research Division

Yes. Yes, sure. So recently you introduced a product that's aimed at high-capacity scale-out environments, much more cost effective, helps address the barricade that flash is just too expensive to use in massive scale. Can you talk about initial response from customers. Does it get you into either a different market or a larger portion of the market?

Lance L. Smith

Well, the product is ioScale. And ioScale was optimized for cost, where ioDrive is optimized for performance. And when you look at a data center, it's -- generally, I mean, you can look at the data center and there's hot data, the stuff that's almost -- that's being ingested real-time. And people want to make use of that data. They want to search on what they want to use it. They want to load up their videos. They want to load up their pictures, and then somebody else wants to go look at it, the stuff that's being immediately used. That's where ioDrive fits extremely well. So it's mission-critical, highest possible performance, ioDrive is a perfect fit. As you move from this hot data all the way to, let's say, archival, right, or something that people would term as cold storage, the data starts cooling off but it's not as needed as much or as often, then the needs or the requirements for those products going into the middle of this data center changes. This is where ioScale fits well. So this is not a cannibalization story. This is ioScale incrementally adding to our opportunity and our market availability. But it also deals with the just good enoughs. Right? The just good enoughs are what's fitting down in this sort of lower cost, lower performance area. But what we've with ioScale, we've really truly tuned it and optimized it for highest capacity and lowest cost. It's a single controller design. It's one PCB, and it is essentially twice the amount of NAND that we would have normally hooked up to a single controller to fit like ioDrive. So you amortize the cost of the infrastructure, the controller, all the raw components, the print circuit board against twice the amount of flash. So an ioScale holds 3.2 terabytes of user space. It's got 4 terabytes of raw and NAND flash on it, where like an ioDrive, the equivalent would be about half as much. It's like 1.3 terabytes worth of capacity -- usable capacity.

Kathryn L. Huberty - Morgan Stanley, Research Division

Right. And how has getting a cheaper product into the market changed your discussions with the big hyperscale companies?

Lance L. Smith

It's opened up opportunities. Quite frankly, it was their idea. They told us what they wanted. And for us, we're able to take our technology since we own all aspects, the controller, the software, the VSL that I described, and it's field programmable. I mean, this is something that we took the input and immediately turned around and created product specific for hyperscale. So we know it fits what they needed because it's what they asked for.

Kathryn L. Huberty - Morgan Stanley, Research Division

Okay. And when you think about ioScale, it sounds like you don't expect companies who were buying ioDrive to now instead go buy ioScale for the most part. Would you expect ioScale companies to add on as much software capabilities, some of the ION software, as an example? Or would you expect it to be truly a lower-cost, lower-margin solution?

Lance L. Smith

So I -- okay, there's a couple of questions. So the first one is that I think that ioScale expands our footprint in the data center. That's very key. When it's mission-critical, most likely it will stick to ioDrive. So someone who's buying ioDrive before and are looking to place it in other aspects of their data usage that requires more capacity at a lower cost, then ioScale will be a perfect fit there. What's nice about that is that now we don't have to take this very high-end, performance-driven design and massively discount it to get it to fit in these other tiers. ioScale is now a good fit there. So we can go and compete against the just good enoughs without having to hurt our business model and our gross margin goals. And for the other aspect about -- in terms of...

Kathryn L. Huberty - Morgan Stanley, Research Division

The software add-ons.

Lance L. Smith

The software add-ons and their functions. Again, our footprint is just getting bigger. So there is no incompatibility between ioDrive and ioScale. All the software we make for ioDrive, the same software works for ioScale. So everything we talked about ION, VSL, directCache, ioTurbine, it all works across the board. It's just a different economic model for the customer to make use of. So we still get that stickiness.

Kathryn L. Huberty - Morgan Stanley, Research Division

And Dennis, maybe we'll just stop there for a second and talk about what you expect the margin characteristics to look like for ioScale versus ioDrive.

Dennis P. Wolf

Sure. So -- and I think I'm going to take it broadly. The -- where we think we get greatest market capture at least at this point in the market is about a 56% to 58% target margin range, ioScale fits within that. ioDrive is above that now. And so therefore, last quarter, as an example, we had a 62% gross margin, and we've been -- we gave visibility for Q3 and 4 that it will be 58% to 60% because we think that there will be more ioScale in Q3 and 4 than there has been. But I think it will still stay within the range that we're referring.

Kathryn L. Huberty - Morgan Stanley, Research Division

Okay, okay. Shifting to more specifics around the software. What type of uptake are you seeing on the add-on software, like ION, just generally speaking. And then, Dennis, at what point would we, as investors, focus on the financial model, expected model of software as a separate line item. When is it significant enough that it becomes important to the margin story and revenue story?

Dennis P. Wolf

Yes, it's not significant enough yet. I mean, we have had good traction with ION, which is a software solution, and ioTurbine. We're beginning to get that traction. It's still de minimis to the model. I know that when we went public 7 or 8 quarters ago, we had talked about the fact that we'd probably be no more than a couple of percent this year, and that's probably what it is. We could double to that number a year from now. To get to 20% sustainable software rate, I think it's probably still a few years out.

Kathryn L. Huberty - Morgan Stanley, Research Division

Okay. Shifting to -- back to the hyperscale, the strategic accounts, because this is a topic on the last conference call, Dennis. You talked about your 2 largest customers moving from $50 million, $60 million quarterly run rate down to a $10 million to $20 million quarterly run rate. Can you just talk about what you think is causing a simultaneous decline in those 2 accounts?

Dennis P. Wolf

Sure. I think I have to take it specific to each one and then maybe blend in the full question. When you look at Facebook -- and we're able to speak more about Facebook. Apple is a very opaque company. We really aren't going to be saying much about it. But when you look at Facebook, the first thing you see when you have a new customer is what they are -- if you're going to be responsible for the infrastructure, you get a really good sense of what that infrastructure will entail. And you're going to have a sense of when the first data center is coming up, it might be simultaneous with the second one. And we have that good traction, right, for a couple of years where we knew that these 3 data centers were going to come up, the fourth data center is probably in this calendar year. What happened though in the process is that during the last -- at least in the last quarter, we found that the first data center that we built was being repurposed for a different tiering level, which we are in. We won, obviously, because they were able to take some of our older product and put it there. And so what we expect is kind of a timing thing. What we expect is for that tier, it should probably start to move up within the next couple of quarters, and then there's that data center that we expect as well. So we would anticipate that our 10% to 20% sustainable run rate will go up from there. As it pertains to Apple, Apple is absolutely -- it grows as their applications grow. So if you're into technology for an i-X product, if that product or application is going to grow, we will grow with it. We anticipate that coming. But with Apple, we really don't get very much visibility. It feels simultaneous. It is coincidental but it is exactly what happens.

Kathryn L. Huberty - Morgan Stanley, Research Division

Okay. And when you think about visibility, you mentioned a few things: one, new data center openings, which you know will come; the penetration within the accounts. And then at Apple, if there was some new application platform or the user base grows, do you think that you'll grow along with that?

Dennis P. Wolf

We will grow with it. And then it also takes you to the next question that you may ask. Behind those -- and actually, Katy wrote an excellent report about it. If Apple and Facebook represent 10% of potential ioScale customers out there, there's still 90% of nurturing that we can aspire to, to be able to grab some of that opportunity. And it's starting to show. I mean, we don't have anything to speak about now except that we think that we're getting good traction with other customers who would be able to provide to us something that would be larger than just a core business, something more like a tier [ph].

Kathryn L. Huberty - Morgan Stanley, Research Division

Yes. On that point, in the December quarter, you had 5 companies that spent over $1 million on a product that you hadn't even announced to the market. I mean, that's pretty unbelievable. And I imagine, 2 were the obvious, and then there were 3 others. Of those other 3, I mean, could any of them become a Facebook-sized customer over time? Is that possible?

Dennis P. Wolf

I think -- no, I don't -- the ones that we did get are really more like large core customers. Right? So when we would -- we've had about 50 customers in excess of $1 million over the last 4 quarters, 5 quarters; 10 last quarter, 5 were hyperscale. We probably wouldn't have been in those customers, 3 of them were telco, a couple were international. Had we not actually shifted gears and moved to an ioScale product, that doesn't mean that they can't be large. But to have a couple of 100 million in a couple of years, there are only a certain number of those customers. They're not in that number yet. And by the way, for -- when we articulated what we expected for the second half of this year at Q3 and 4, we also didn't layer into those numbers any expectation for some of those strategic that could come.

Kathryn L. Huberty - Morgan Stanley, Research Division

So your current guidance over the course of the next few quarters does not assume any new strategic customers?

Dennis P. Wolf

That's correct. For the next [ph] couple of quarters.

Kathryn L. Huberty - Morgan Stanley, Research Division

Okay, okay. Good. Shifting to the core business. The growth rate with 67% year-on-year core, so this would be enterprise customers and some hyperscale going through your OEM partners, grew 67%. You're expecting a healthy 50% growth over the next couple of quarters. Talk about what's driving that growth rate and how do you sustain that growth rate.

Dennis P. Wolf

So that growth rate is actually -- look at it in 2 different ways. We have like 4,000 customers. But of those customers, the vast majority of those customers are 2 or 4 core kind of customers. What we're finding though now is that the order sizes are getting larger. So the pipeline is getting thicker, and the actual deal itself is getting larger. So we think that, that's part of what's happening in core. The other is just scale. I mean, we now have over a couple of hundreds salespeople, 100 AAEs and about 100 SCs, and they're scaling fast. And they're expanding market. Asia is a good example of what's happening. If you look at last quarter, roughly 40% of our business is either a -- had a home address at either Asia or Europe. So that core build-out is occurring. We're 19 countries now. I think also, the fact that we have a hyperscale product to talk about now, it gives us even more opportunities that are opening, just like those 5 that we talked about just a minute ago.

Kathryn L. Huberty - Morgan Stanley, Research Division

Okay. And what are your expectations for the CISCO and NetApp relationship as it relates to growing the core business?

Dennis P. Wolf

Yes. So we've been working with them for a couple of quarters now. I think it's -- we predicted that in the first half, it wouldn't be as significant as the second half. So we expect more business from Cisco and NetApp going into the second half of the year and next year. It's meeting what we had expected.

Kathryn L. Huberty - Morgan Stanley, Research Division

Okay. And would you -- just a follow-up on that, with a technology like this, it takes somebody to come into an account and evangelize it. Are you expecting that partners, like Cisco and NetApp, become the evangelist? Or do you still have to get into those accounts and your own salespeople have to send that message?

Dennis P. Wolf

Both. I mean, I think you get evangelizing after they get familiar. So we've trained like 500 sales people. And Cisco, obviously, has a faster path to it because it's new for them. NetApp, it's really -- we're working with people who are very well-versed in what we do. And so it's really a matter of time. I think the evangelizing is going to come soon.

Kathryn L. Huberty - Morgan Stanley, Research Division

Okay, okay. Let me follow-up one question on gross margins, and then we'll see if there's any questions in the audience. Dennis, you talked about the 56% to 58%. You're running ahead of that on ioDrive right now, but that's the long-term target where you optimized growth. One of the obvious component input is NAND pricing, and there is some concern in the market that NAND pricing isn't dropping at the rate that it was 12 months ago. And so how do you think about the current trend of NAND pricing influencing gross margins? And do you have contracts that smoothed that out over time?

Dennis P. Wolf

So yes, we do have contracts, and we typically have 6 to 9 months of availability out there. We're -- so we predicted, we've been saying this year that it would be relatively stable pricing for NAND. And for us, it has been even though the market isn't necessarily that. We try to layer on top of each contract to rolling other contract. We work with primarily 3 key vendors, and we'll continue to do that. I'll also say though that if things go completely out of hand in NAND pricing, that it will have an effect on us. About every 10% shift in price is about a 200- to 250-basis-point gross margin effect for us. But right now, we're still anticipating, and I'll let you expand on it since you've had all the time. But we still anticipate having a relatively constant price.

Lance L. Smith

I would only reiterate that we have a really good relationship with our NAND suppliers, and it goes beyond just a buy-sell relationship. We do a technology exchange between them. We have the knowledge of how that NAND is being utilized. And it's taken favorably with these suppliers, plus we've always been a good buyer. We're very consistent about our NAND purchases. We don't go out in the open market to try and purchase NAND that's just available on the spot market. We work with the direct suppliers that gives us consistency of supplying us, gives the consistency of quality, and we really do hold those relationships very strong and very close with them. So right now, it actually is in pretty reasonable shape.

Kathryn L. Huberty - Morgan Stanley, Research Division

Okay. And it's sounds like you have maybe 6 months visibility, and we'll have to watch NAND pricing from there. Okay. All right. Let me stop right there and see if there's any questions from the audience. Here in the front?

Unknown Analyst

Maybe if you would just expand a little bit on some of the success and/or interest levels you guys have internationally. I know you have relationships, I think, with China Mobile and Baidu, maybe outside of China, other areas that you're spending time on potential big opportunities. And then maybe just one more time to kind of reiterate, where are the comp ends coming from, from an investor standpoint on fiscal '14 and Apple and Facebook coming back strongly?

Dennis P. Wolf

Okay. Why don't we take the -- you take first, which is the geographic expansion.

Lance L. Smith

So geographically, we've always said, since the IPO and the S1, we said that the majority of our folks will be expanding outside of North America. We spent quite a bit of time in Europe, in the EMEAs and in China and Asia Pac. In China, particularly, we've had a very big focus there because -- for obvious reasons, the population size, the demand and, in fact, the growth in the Internet and online commerce. That continues to grow. In fact, that's actually doing quite well when I think about 130% growth from quarter-on-quarter in China alone. In the EMEAs, we've seen the POs just in EMEA double over the last quarter. So we continue to invest there by placing more AAEs, placing more SCs and support staff so that we can grow out those businesses themselves and instruct the folks on varied given verticals and applications. I think we're just really at the start of that effort and starting to see some fruit being beared from those opportunities.

Dennis P. Wolf

As it pertains to Facebook and Apple, the question was visibility, primarily. The visibility that we do have is, again, to the tier that we're winning, that our product has just moved into, as well as future data centers. So we still have a good sense of their architecture at Facebook. And again, with Apple, it's really application-specific. What we do know about this business is that this business that we have requires more than strategic. We know that this business is about core growth, and we know it's about taking the successes that we've had in Apple and Facebook and replicating them amongst up to 25 potential companies that could have the scale capability that Facebook and Apple has. So to me -- and we're not talking '14 yet, but to me, the success path is to continue to grow out that core just as quickly as possible, while at the same time, winning a few of the other potential strategic accounts in 2014.

Kathryn L. Huberty - Morgan Stanley, Research Division

We have about a minute left. Any other last questions from the audience? No? Okay. We'll wrap it there. Thank you very much.

Dennis P. Wolf

Thank you very much.

Lance L. Smith

Thank you very much.

Copyright policy: All transcripts on this site are the copyright of Seeking Alpha. However, we view them as an important resource for bloggers and journalists, and are excited to contribute to the democratization of financial information on the Internet. (Until now investors have had to pay thousands of dollars in subscription fees for transcripts.) So our reproduction policy is as follows: You may quote up to 400 words of any transcript on the condition that you attribute the transcript to Seeking Alpha and either link to the original transcript or to www.SeekingAlpha.com. All other use is prohibited.

THE INFORMATION CONTAINED HERE IS A TEXTUAL REPRESENTATION OF THE APPLICABLE COMPANY'S CONFERENCE CALL, CONFERENCE PRESENTATION OR OTHER AUDIO PRESENTATION, AND WHILE EFFORTS ARE MADE TO PROVIDE AN ACCURATE TRANSCRIPTION, THERE MAY BE MATERIAL ERRORS, OMISSIONS, OR INACCURACIES IN THE REPORTING OF THE SUBSTANCE OF THE AUDIO PRESENTATIONS. IN NO WAY DOES SEEKING ALPHA ASSUME ANY RESPONSIBILITY FOR ANY INVESTMENT OR OTHER DECISIONS MADE BASED UPON THE INFORMATION PROVIDED ON THIS WEB SITE OR IN ANY TRANSCRIPT. USERS ARE ADVISED TO REVIEW THE APPLICABLE COMPANY'S AUDIO PRESENTATION ITSELF AND THE APPLICABLE COMPANY'S SEC FILINGS BEFORE MAKING ANY INVESTMENT OR OTHER DECISIONS.

If you have any additional questions about our online transcripts, please contact us at: transcripts@seekingalpha.com. Thank you!

Source: Fusion-io's Management Presents at Morgan Stanley Technology, Media & Telecom Conference (Transcript)
This Transcript
All Transcripts