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

Executives

Peter J. Ungaro - Chief Executive Officer, President and Director

Analysts

Jerry Liu - Morgan Stanley, Research Division

Cray Inc (CRAY) Morgan Stanley Technology, Media & Telecom Conference February 25, 2013 4:55 PM ET

Jerry Liu - Morgan Stanley, Research Division

Hey, everyone. Let's get started. My name is Jerry Liu. I'm with the Tech Hardware team here at Morgan Stanley. And it's my pleasure to welcome Peter Ungaro and Cray to our session right now. First, let me just say, note that all important disclosures including personal holding disclosures and Morgan Stanley disclosures appear on the Morgan Stanley public website at www.morganstanley.com/researchdisclosures or at the registration desk. With that out of the way, Peter, welcome.

Peter J. Ungaro

Thanks, Jerry.

Jerry Liu - Morgan Stanley, Research Division

So maybe you can just give us a quick overview of Cray for anyone that's not familiar with sort of the things that you have going on recently, talk about the key products and the end markets.

Peter J. Ungaro

Okay, great. So thanks, everybody, for joining us here. The -- Cray has been around for a long time. We've been a broad name in the world of supercomputing since the '70s. And Cray has really been a company that has been, at our core, an engineering company, developing new technologies really for delivery into the supercomputer market space. And we've been a leader in that space around the world today. In fact, at the very high end of the market, it's Cray and IBM is pretty much over 80% of the worldwide high-end supercomputing marketplace today. But over the last couple of years, where we've been really investing some time and energy is to take these supercomputing technologies and start to move them into new markets, so markets that are -- have faster growth, have potentially higher returns in those markets and predominantly around the world of Big Data and storage. That's been a big effort overall. We've first came out with a new storage product here, just about a year ago -- just under a year ago, I should say, and that's been going very well for us, as well as with a new product in the Big Data area, graph analytics space, which I'm sure we'll talk a little bit about, that we just made generally available here at the start of the fourth quarter of last year. So some really kind of a company that's been, I think, a historic leader in our space, and now, really starting to expand where we can take that into some adjacent markets over the last year.

Question-and-Answer Session

Jerry Liu - Morgan Stanley, Research Division

Yes. No, I definitely want to touch on some of the growth areas, but starting with the core business to get an understanding there. What are some of the key drivers? Public budgets is definitely a big factor. Do you think data growth or any -- what are the key drivers you look at for the traditional high-end supercomputing?

Peter J. Ungaro

Yes, the high-end market is really been driven off 2 things, I would say, overall. So the first is just the need to do better simulations. So today, what you use supercomputers for it to represent reality. So instead of doing physical prototyping, for instance, if you're a manufacturing company or if you're an energy company, instead of laying down a drill, you want to simulate all of that as much as you can within a supercomputer before you start to invest that dollars into building your first prototype. So it's a big driver overall. The same even in national security, it's much better to try and simulate environment than it is to go and try to secure and control all those different control points. The other piece is that the market overall is really bifurcating, and it's bifurcating to what I like to call a cloud or a Cray. It's bifurcating either into a very commoditized, very open, driven by costs environment that we typically find in cloud infrastructures today or very tightly-integrated environments where you can get very, very high performance, much different kind of performance characteristics than you can get in a cloud. And this is really where Cray is very focused on, is this very tightly-integrated environment.

Jerry Liu - Morgan Stanley, Research Division

I'm glad you brought up the sort of cloud or the scale out kind of competition. I mean, how do you see that playing out? Is that -- I mean, that does have, obviously, some competitive risk to some extent. But do you feel there are just things that they can't do that you guys can do? How do you compete with them?

Peter J. Ungaro

Yes. I really -- I see it as more complementary than competitive, actually. Because today, for applications that can run in a cloud, it's going to run in a cloud at way lower costs than it's going to run in a Cray-like environment, and so not very competitive from that perspective. But because of that, and the way that clouds are architected, very loosely coupled machines, you can't do some of the types of applications that people would run on their Cray. So for instance, most people like to use hundreds or thousands of processors on a single application when they're running on a Cray system. So you need very tight coordination amongst all the processors, very high bandwidth between all the processors, memory and processors, and through your system interconnect in a very integrated software environment that you just can't get in a cloud environment today. So I see them as very complementary environment actually, and it allows us to feel very differentiated machines because you don't have to worry about this kind of processing because you can do it so much more cheaply than you can on a tightly-integrated environment.

Jerry Liu - Morgan Stanley, Research Division

I see. Maybe we should just touch on your flagship product here, the XC30. When you look at sort of the current product cycle, how does that compare to historical trends in terms of pipeline or revenue generation or margin at this point?

Peter J. Ungaro

Yes. So the XC30, we just started shipping in December. I couldn't be more excited about a product than this one. This is our biggest R&D project that we've ever done before as a company, and we just started shipping it. Our products typically have about a 4- to 6-year life. So we're typically bringing out a product like the XC30, and then we'll go through multiple upgrades over a period of time. And before we started doing this, we used to just have a one at a time product and it had a much shorter product life. What you'd see is about 80% -- 70%, 80% of the revenues or the margins that you get out of this would be in the first like 9 months of the product life, and then it will quickly dwindle down. Well, because we've been able to upgrade these last -- and we've been through 2 generations, this will be our third generation of product where we can continue to upgrade and drive total cost of ownership down through a cycle, we've actually gotten higher revenues in the last year of the product life than in the first year. And by the way, higher gross margins, too. So it's been a really interesting plan for us where we've been able to continue to grow throughout the life cycle, all the way through to the end.

Jerry Liu - Morgan Stanley, Research Division

So the sweet spot for the XC30 is kind of yet to come?

Peter J. Ungaro

I think it's yet to come, yes.

Jerry Liu - Morgan Stanley, Research Division

Okay. And just another thing on the core supercomputing market, and that's in terms of the end market exposure, over -- I think, about 90% of that is exposed to either government labs or educational labs. And obviously, there's a lot of concern about budgets in the last couple of years. So how does 2013 compare to the last couple of years? How do you see 2013 playing out?

Peter J. Ungaro

So I think in the U.S., so if we look at the U.S., because while we're about to go into sequestration in the U.S., I will tell you, Japan is going through a huge investment phase. So when you look at it on a worldwide basis, it does more balance out. And when you talk about the 80%, 90%, that's really on a worldwide basis. But in the U.S., I would say supercomputing and high-performance computing, as well as Big Data and storage, have gotten a really great track record, I think, in the budget crunches. Because typically, people are applying dollars there because they can do it much more cost effectively than building a new instrument or building a new facility to demonstrate something. So we've done really well. In our current outlook for '13, we've kind of knew sequestration was likely, so we've kind of factored some of that in overall to our plan, so we don't feel it's going to have a big impact to us in '13. And we'll just have to really see how that plays out over the next few years.

Jerry Liu - Morgan Stanley, Research Division

And then really outside of the U.S., you wouldn't really take on any specific region or country where there are concerns?

Peter J. Ungaro

I think that there -- it's country by country. As I mentioned, in Japan, they're doing a stimulus right now, so it's kind of at the opposite end of the spectrum. In certain countries in Europe, some are struggling, some are doing very well. So in -- but in the bigger countries where we typically have been selling our systems, like in Germany and the U.K. and such, they seem to be very strong right now, and Switzerland, very strong.

Jerry Liu - Morgan Stanley, Research Division

Okay, okay. And you brought up sort of your planning, your target for 2013. You look at sort of $500 million revenue target. You said 45% of that will be the December quarter. So the question then is, what kind of visibility do you typically have? And is this -- like what's the...

Peter J. Ungaro

You're asking hard questions. So we are expecting good growth. We've had a few years now of really great growth, and we expect 2013 to be another year of really nice growth. And it is about, right now, our current outlook is $570 million in the first quarter, kind of growing quarter-over-quarter, and about 45% of the year in the fourth quarter. And a lot of that is because in our high-end business, we have tremendous visibility. It's not perfect, but we typically can see our pipeline. We're typically building those systems quite early in their life cycle. And so we have a really good visibility in through the year. So I feel like we're within a nice balance in that business, and that's our biggest business. The other businesses, both our cluster solutions business, which we just acquired from Appro here at the end of 2012, as well as our Big Data and our storage businesses, those are businesses that have been growing kind of quarter on quarter on quarter. And so when you stack up 3 businesses that have been growing quarter-over-quarter, and we see as we look at the pipeline right now that we think that's going to continue, it just starts making for bigger quarters the further you go out. So it's really a combination of those 2 things: good visibility in our core business, and we kind of see that a lot of that is based on a next-generation processor from Intel called Ivy Bridge, which we expect to be big in the second half of the year; and then, these growth businesses that are going quarter-over-quarter.

Jerry Liu - Morgan Stanley, Research Division

Okay. I'm going to ask another question on sales -- on the sales force, then we're going to shift to the growth topics. And before we shift there, if you guys have any questions, I'll pause a couple of times during this session, so get your questions ready. But on the sales force, last year, you grew it by 50%. So I'm just wondering, when you look at the longer-term strategy, what is the reason that you step it up so much, and what do you want to do going forward for the next year or 2?

Peter J. Ungaro

Yes. As we look at kind of the past few years at Cray, a lot of our investment has really been in R&D, and it's been building out the kind of a really differentiated product portfolio, not just in our high-end machines, but doing the storage product, our Big Data offering, et cetera. As we look out going further, we really see our growth investment shifting more to the go-to-market side of the business, and we really started that in '12, where we started to add in a lot into our sales force, marketing, presales, all those sides. And I think that that's going to have to continue as we grow the business. So I think that what you'll see is incremental investments starting to shift a lot more into that side, the sales and marketing side of our business, while we do a lot of R&D in our company. But I think that, that can be really held where it is today and leveraged upon a lot more. So I do see that continuing to happen. We're doing that both in industry verticals, so looking at new verticals for ourselves, as well as geographic coverage.

Jerry Liu - Morgan Stanley, Research Division

I see. Let's just pause a quick moment and see if there are any questions.

Unknown Analyst

I do have one question on the...

Jerry Liu - Morgan Stanley, Research Division

We'll give you the mic right here.

Unknown Analyst

It's my understanding that GPUs are kind of inherently parallel-based, much like HPC. And so my question is kind of in 2 parts. Do you kind of use a GPU-CPU mix in your XC30 or some of your new machines? And then secondly, with regards to that bifurcation that you just talked about, do you see a potential threat from the -- if someone were to build a cloud-based GPU-CPU offering and wire together with InfiniBand or something, could they potentially offer kind of a cloud-based, more HPC, kind of invade your niche from there?

Peter J. Ungaro

Okay, great questions. We actually have the biggest GPU-based machine in the world today, so we've done a lot of work. And what we found is GPUs are inherently parallel, as you mentioned, very high-performance at the node level, so at the individual GPU or processor level. And what that ends up doing is it puts -- because you have so much processing on the node, now it puts more pressure on how to feed that node, how to keep that node fed with data. And that adds a lot more pressure to the parts of the system that we differentiate on. So it actually improves our differentiation as that node gets more powerful. And so we're very excited about GPU computing. Intel has a project called many-core computing, Many Integrated Core, as they call it, and they're file line. And so it's very -- to us, it's a very exciting trend overall because it allows us to -- actually our differentiation, to become more important in the core machine than it even is today, which is great for us. Can somebody build that kind of service on a cloud and stuff? Yes, but typically in order to get the cloud economics, so in order for that machine to be economical over a cloud, you have to architect the machine very differently than you'd architect it to be what we sometimes call a private cloud, right, where you can have an own integrated Cray machine all optimized for that one platform. Because they have to optimize for a broader set of applications and such, we don't think that that's going to be very competitive to what we do, but it can take a big part of the overall commodity cluster marketplace. Again, not really where we play so much, but it can take a part of that market. And that's why I do see this bifurcation, I see a lot of the energy right now is in the middle of the market where you have kind of mid-range systems that are InfiniBand-based clusters and stuff. And I actually see a lot of that either going -- migrating towards the cloud or migrating up into these more tightly coupled machines like we're building. And so that's that bifurcation that you see.

Unknown Analyst

And just could you give the estimate of how much government and academic is the -- for your niche? Did you say 80% or was I mixing that up?

Peter J. Ungaro

No. Overall, it's -- our business right now is about -- is up to 10% commercial, 90% is through either federally-funded programs -- this is worldwide, not just in the U.S., of course, state government, different governments systems, sort of academia or research laboratories, whether they be public or private. So that's about 90% of our business, commercial is about 10% of our business. That's actually a big shift. A couple of years ago, commercial was about 2% of our business. And so we've really started to grow that side, and that's been actually a really exciting part of our growth story overall, yes.

Jerry Liu - Morgan Stanley, Research Division

Any other questions?

Unknown Analyst

For the computer simulation market and special commercial, maybe just some deeper thoughts or broader strokes on where you think we are as far as early innings, and what's driving that ecosystem, how you guys think you can compete and win, and who compete against for those sort of systems.

Peter J. Ungaro

Sure. The -- I think that we're kind of like in the second game. So in the '80s, there was a lot of work going on in industry around using computer simulation. So Boeing started using computers to simulate jetliners, and we were crashing cars with computers instead of prototypes and all that. So that -- and really the whole seismic analysis piece of the petroleum market really started. So that happened, and they were using very high-end supercomputers, and then that really commoditized. As the kind of crunch came on companies, that really went down into commodity and even into cloud environments today. And what we're seeing is a resurgence in that space, and we're seeing the resurgence really because we're able to do so much more now in computing. So for instance, in seismic processing, new algorithms that they can really view the subsurface of the earth in a much better way than they could before. All of a sudden, it's starting to put demands back on the computers that they hadn't had before, and so they need to move up to better supercomputers. Instead of simulating, for instance, we have a major manufacturing customer that was simulating -- of a jet engine, they were simulating one fan blade, basically, at the time of the jet engine. Now they can start to simulate entire engines on their -- on a single supercomputer. Boeing uses a Cray supercomputer for a wing of a Dreamliner jet to look at the entire wing, airflow over the wing, materials within that wing, et cetera. So I think we're really starting to see a resurgence in this area where the models and the sophistication of the engineering is coming to a point where it's starting to need that kind of power again. And that's where we've seen our supercomputers really get traction, is not in kind of the traditional space, but in kind of these new spaces that are much more sophisticated analysis than we had in the past.

Unknown Analyst

And it sounds like that's like a, I'm not sure what the right phrase is, multi-varied or multi-function sort of simulation where you're doing a lot of -- you're simulating a lot of things in parallel at the same time. Is that kind of the main point?

Peter J. Ungaro

Exactly. Different materials as well as airflows, and so you're combining different types of processing all in one single integrated model, yes.

Jerry Liu - Morgan Stanley, Research Division

Question in the back?

Unknown Analyst

Just as a follow-up to that question. So in the '80s, you started out that way, you did really well, and then that became commoditized. So what's -- how is this era different? I mean, what's going to prevent competitors to, maybe in a couple, 3, 5 years, to catch up and come out and take the margin out of the business?

Peter J. Ungaro

Yes. So I think it's this bifurcation that's happening. So the interesting thing in the supercomputing market, in the '80s and '90s especially, and even in early 2000s, huge increased competition in the space as more and more people got interested in that. I was at IBM at the time and restarted kind of supercomputing effort inside of IBM right now, that's been going on now and it's a multibillion business for them. So it's -- at that time, there's a lot more interest in that because it was a big growth part of the market. As time has now shifted, much more of the growth is in the cloud piece of the market. And that's where a huge part of the growth is, and actually declining competition in the high-end. And now you get Cray and IBM jointly owning probably 80% of the worldwide market of high-end supercomputers, really the only 2 companies I would argue doing significant amount of research in this area right now. So I feel like we're in a pretty good space, and we keep on improving on that, so I think we keep on leapfrogging, which means we're building kind of a barrier into that market over time that I feel pretty confident in.

Unknown Analyst

I apologize for the naivety of the question, but why is the fact that it's the cloud this time make it different and less penetrable?

Peter J. Ungaro

So really, it's not that it's the cloud that's making it different. I mean, the cloud is giving people a different way to access things at much lower cost, and it's basically taking -- their -- the cloud is commoditizing commodity, if you could think about it that way, and really taking that low-cost part of the market. And so what's left, I think, to innovate around is the parts of the market that the cloud really can't address because of the way that you have to architect systems on a cloud. And that's where companies like Cray can be much more specialists and play to that part of the market where there's not a lot of players doing today, because there's not as high of a volume, not as big of a market. The entire supercomputing market worldwide from systems that are $0.5 million on up is only $4 billion. Now $4 billion for a company of Cray's size is a gigantic cam, right, a huge addressable market. For a company that's IBM size, that is a very small niche market. And so it really allows, I think, companies like us to play in a specific market and really differentiate in that space.

Unknown Analyst

What's going on with budgets in the research area in universities? Is that potentially a risk for you guys?

Peter J. Ungaro

I think a lot of it is around kind of the risk of sequestration. Because so far, research budgets have been really strong. And in fact, research budgets tied to supercomputing has been even stronger. So I feel that that's been really great. And what we're going to have to see is how does the government in the U.S. handle sequestration? What do they do with the budgets, and how does that all play out? I think that -- I think it's going to get cut over time slightly, but I really don't think that it's going to harm Cray so much because we're building such a differentiated offering in that market. I think people will reduce spending in the mid-range and lower tiers before they stop spending in the upper tiers because that's where they're getting the biggest breakthroughs, that's where they're making their names, that's where a lot of their energy and focus is. So I think we're in a pretty good spot, but it's impossible to predict exactly how the government is going to handle the current situation. I couldn't attempt to that -- do that.

Unknown Analyst

Can you talk about your pipeline a bit longer term out? For example, in this commercial simulation market, once the GEs or the sort of Boeings have bought a supercomputer, they're probably fine for the next 10 years. And so this is my thinking before. How much visibility do you have, a few years out for that?

Peter J. Ungaro

Typically, we see people upgrade their systems every 2 to 3 years, because the technology moves so far, and that's one of the reasons that we have enabled our platforms to be upgraded in place so we can leverage a huge part of that infrastructure and do low-cost upgrades over time. And that really actually drives down the cost of computing for customers. So you end up in a situation where technology is moving at such a fast pace that you can actually upgrade and lower your total cost of ownership as you do upgrades over time. So typically, we're seeing 2- or 3-year cycles with the technology, not 10 years, even in commercial.

Unknown Analyst

Can you give us some use cases of how your products are used in Big Data world?

Peter J. Ungaro

Big Data? So our first Big Data offering is really around graph analytics. It's a product that we call uRiKA, and it's a -- graph analytics is a space of the market which is pretty unique. So mostly in Big Data, companies today are doing search, so you -- needle in a haystack. You have a lot of data, a haystack, and you know what you want to find, and you search for that one piece, that needle, and pull that out of the haystack. What we're really focused on with graph analytics in our uRiKA product is discovery, where we don't really know what we're looking for in the data. We're trying to find something that's unknown or hidden in a bunch of related data. And so what we've done is we've built a platform where we can analyze the data, test hypotheses, so make a hypothesis, test it out quickly, use that data to create another one and continue to test out multiple hypotheses until you find that unique or hidden piece of data that's typically very high value. So one place that we've started to do a number of engagements is around the area of cyber security, so threat analysis. Threat analysis is a very challenging area of cyber security, where you have -- you're not sure what's going on. You have these unusual things in your network and you're not quite sure exactly what it is and you want to look at all the data, all the data off your firewalls and your DNS server, and you want to map that against known issues and security threats, and you want to bring all that data together and really to find and solve a process. And for instance, our team that's doing this, actually a couple of them worked at a cyber security center in the government where they have potentially 100 analysts. And they would typically have -- on a Monday morning, they would find kind of a number of hypotheses that they want to test, they'd have teams of 4 to 6 analysts go out and test these. And within a day or 2, they bring back an answer. With this machine, we're able to do this in a course of a few minutes or even seconds. And so it's a very exciting kind of way to really test hypotheses very, very quickly. So cyber security is one, another is news analytics. So looking at what's going on in the news, looking at how do we do trading or how do we find -- how is this going to impact another part of a company's business, or in the financial services industry, how a news event in one part of the world trigger maybe trades that you'd want to make, web-based [ph] trading and things like that. And then we're doing another area around life sciences where we're doing genetics basically, competition [ph] of biology, where we're looking at cancer genomics and really trying to look at the correlation between all these known entities to help as find cures for cancer.

Unknown Analyst

[Indiscernible] how big this market is? Or how do you guys think about the size of the market, the potential opportunity?

Peter J. Ungaro

Yes. The overall broad Big Data market is hard to grab a hold of right now. If you look at the different analysts, everybody views it differently. I would say, on average, it's maybe about $2 billion. But this graph analytics space is pretty nascent part of the market, it's brand new. So we think it's going to grow very, very fast. Big Data overall is growing at 40%. I think this part of the market is going to grow even faster than that. And there's been some really exciting things in the past couple of months that have happened in the graph analytics space, probably the biggest being Facebook announcing Graph Search, right, as being a real application for using graph and doing that on such a public entity like Facebook.

Unknown Analyst

Sorry, I walked in late, but -- so your business model is still to sell hardware primarily, correct? Or is there something in addition to that, that I'm missing?

Peter J. Ungaro

Well, it's shifting over time. So we always -- we think of our systems as solutions. And typically, we've invested in hardware. But over the past few years, the majority of our R&D investment has been in software, much more than hardware. And as we've gone to market with our new storage offering with Sonexion, that's really more of an appliance-size offering, right? Hardware, software, service, all integrated in one offering. uRiKA, hardware, software, service, all integrated in one common offering. So more and more, we're getting this really integrated view of our business, not just hardware, but hardware, software, services, all kind of coming together in much more of an appliance-size model over time.

Jerry Liu - Morgan Stanley, Research Division

Let's take the rest of the questions offline. I think we've really run out of time. So, Peter, I thank you very much. And thanks, everyone, for joining.

Peter J. Ungaro

Thank you. Thanks, everybody.

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: Cray's CEO Presents at Morgan Stanley Technology, Media & Telecom Conference (Transcript)
This Transcript
All Transcripts