NVIDIA Corporation (NVDA) Q1 2013 Earnings Report, Transcript and Summary

Good morning. My name is Ashley, and I'll be your conference operator today. At this time, I would like to welcome everyone to the NVIDIA Financial Results Conference Call. [Operator Instructions] Thank you. Mr. Christopher Evenden, Senior Director of Investor Relations, you may begin your conference.

Thanks, Ashley. Good morning, and welcome to NVIDIA's conference call on the first quarter of fiscal 2013 results. With me on the call today from NVIDIA are Jen-Hsun Huang, President and Chief Executive Officer; and Karen Burns, Interim Chief Financial Officer. After our prepared remarks, we will open up the call to a question-and-answer session. [Operator Instructions] Before we begin, I'd like to remind you that today's call is being webcast live on NVIDIA's Investor Relations website and is also being recorded. A replay of the conference call will be available via telephone until May 15, 2012, and the webcast will be available for replay until our conference call to discuss our financial results for our second quarter of fiscal 2013. The content of today's conference call is NVIDIA's property and cannot be reproduced or transcribed without our prior written consent. During the course of this call, we will make -- we may make forward-looking statements based on current expectations. These forward-looking statements are subject to a number of significant risks and uncertainties, and our actual results may differ materially. For a discussion of factors that could affect our future financial results and business, please refer to the disclosure in today's earnings release, our Form 10-K for the fiscal year ended January 29, 2012 and the reports we may file from time to time on Form 8-K filed with the Securities and Exchange Commission. All our statements are made as of today, May 11, 2012, based on information available to us as of today. And except as required by law, we assume no obligation to update any such statements. Unless otherwise noted, all references to market research and market share numbers throughout the call come from Mercury Research or Jon Peddie Research. During this call, we may discuss non-GAAP financial measures. You can find a reconciliation of these non-GAAP financial measures to GAAP financial measures in our financial release, which is posted on our website or in the case of our fiscal year 2013 outlook, the reconciliation is posted on our Investor Relations website. With that, let me begin. Each of NVIDIA's businesses performed at least in line with expectations for the quarter with notebook graphics and Tegra coming in slightly higher than we had anticipated. As expected, we faced headwinds from the soft economy and hard, hard drive prices -- high hard drive prices, which passed the PC mix away from the channel in favor of OEMs. However, our competitive position was enhanced in mid-March with the launch of our new Kepler GPU architecture. Kepler is more powerful, more compact and more efficient than anything before it. Kepler is also our first architecture to include virtualization technology built right into the GPU. And we'll talk more about what that means at our GPU Technology Conference next week and at our upcoming Analyst Day on May 24. Our first Kepler product was the GeForce GTX 680, which is 10% faster, 20% smaller and 22% more power efficient than its direct competitor. Of more than 130 media reviews published so far, each and every one of the recommended our GeForce GTX 680 over the competition. Kepler is the most efficient GPU architecture we've ever created and we expect this to translate into more market share and higher margins. Less than 2 weeks ago, we launched our second GeForce Kepler product, the GeForce GTX 690, which uses 2 Kepler GPUs to set a new bar for performance. The 690 is a gorgeous piece of engineering and industrial design, and this is widely noted by the press. The public reaction speaks to its degree of craftmanship and pride our engineers and designers put into their product. PC Magazine, which is a mainstream technology publication, described it as the card to cherish. Enthusiast sites went even further with HardOCP calling it the most perfect card they've ever tested. Their review found it to be, and I quote, "the finest looking, best performing, most efficient jewel GPU card in the world with unsurpassed design and engineering." And yesterday, we launched the GTX 670 to even better reviews. These cards enter a healthy PC gaming market. We launched the 690 at our GeForce LAN/NVIDIA Gaming Festival in Shanghai. The event had 8,000 live attendees competing with a passion that has to be seen to be believed, and more than 1 million people viewed the 2-day contest online. Note that revenue hit a record in the first quarter as we refreshed our notebook GPUs top to bottom with the launch of the GeForce 600M series. The unprecedented power efficiency of GeForce 600M notebook GPUs has enabled new notebook form factors like that of the Acer Aspire Ultrabook M3. Putting a GeForce 600M GPU in an Ultrabook makes the Ultrabook truly worthy of the ultra tag. All indicators are that the strong growth will continue for our notebook business through 2Q -- Q2 and Q3 driven by Kepler design wins on the Ivy Bridge platform. Our visibility into OEM builds suggests that the attach rate, stable for the last 3 years, will remain strong. We'll be instituting more Kepler GPUs in the coming quarters. Demand is high for Kepler and although supply will continue to improve, we are not able to meet all our OEM and channel demand in Q2. We do not expect the 28-nanometer supply situation to resolve itself until later this year. This is factored into our current outlook. At this point, we can sell everything we can get. I'll now turn to our Tegra business. The world's first Tegra 3 phone was launched in February by HTC. The HTC One X is their new flagship product and has delighted reviewers worldwide. This is the first phone we've had that launched day one across multiple carriers. Tegra 3 gives it a long battery life, thanks to the 4-PLUS-1 quad-core architecture, console gaming -- console quality gaming and fast web browsing. And because Tegra 3 utilizes a specially enhanced 40-nanometer process, it's not affected by 28-nanometer supply constraints. Other phone wins around -- announced at Mobile World Congress includes ZTE with the first Tegra plus Icera phones, Fujitsu, LG and K-Touch. Tegra 3 phones are now shipping from 22 carriers in Europe and Asia and more will follow shortly. The launch this past quarter of the first 2 phones with Icera baseband processor signals a new opportunity for NVIDIA to penetrate the mainstream smartphone market and to capture a greater share of the silicon bond in each phone. Looking forward, you'll see Tegra 3 LTE phones with partners' baseband processors in the second half of this year and then next year with our own LTE baseband processors as well. Additionally, we remain very excited about Tegra's opportunity in Windows on ARM. Progress continued this quarter with Microsoft shipping development PCs powered by NVIDIA's Tegra 3 mobile processor to Windows 8 developers and to device makers. Microsoft, itself, has made some important recent announcements about Windows on ARM, particularly those referring to its ability for enterprise application. And related, Forrester has predicted that 375 million tablets will ship in 2016 and almost 1/3 of these will be bought directly by business. Even for the rest, Forrester believes that most consumers will use them for work as well as at home. Before handing things over to Karen, I'd like to remind everyone that our annual GPU Technology Conference takes place next week. We're bringing together several thousands of sciences, engineers and academics together to share, learn and network. Jen-Hsun will be presenting a major keynote at 10:30 a.m. on May 15, and it's well worth listening online if you can't be there in person. We'll also have keynotes from a Princeton researcher using GPUs to model collective behavior and the leader of a team competing for the Lunar X PRIZE. Many of the best minds in computing will be there. Finally, I want to highlight our Annual Investor Day on May 24. If you'd like to attend, please e-mail me or visit our website at nvidia.com/ir to register. Now, over to Karen.

Thanks, Chris. Revenue was at the high end of our outlook at $924.9 million on the strength of our new Kepler GPUs we launched in the quarter, as well as from strong performances in our notebook and Tegra businesses. GAAP gross margin for the quarter was 50.1%. Non-GAAP was 50.4%, both at the high end of our outlook, again on the strength of our new Kepler GPUs and desktop and notebook. Our GAAP and non-GAAP gross margins were 1.3 and 2.1 percentage points lower than the prior quarter, respectively, largely due to the impact of 28-nanometer supply shortage that we experienced for our new Kepler products. GAAP OpEx was $390.5 million, non-GAAP was $348 million, both slightly above our outlook. The quarter-on-quarter increase was largely attributable to planned hiring and related infrastructure spending to support our strategic businesses as well as the impact of the payroll FICA tax reset that affects every first quarter. These results contributed to a GAAP net income of $60.4 million for the quarter or $0.10 per diluted share and non-GAAP net income of $97.5 million or $0.16 per diluted share, both slightly above the midpoint of our expectations. Revenue results by business segment were as follows: Our GPU business was down 6.7% quarter-on-quarter due to the 28-nanometer supply shortage, which largely impacted our high-end desktop business as well as shorted supply for our new Kepler GPUs for notebook. Our professional business was down slightly at 4.2% from the fourth quarter, within normal seasonality. Our Consumer business was up 20.8% on the strength of our Tegra business as we've launched Tegra 3 for smartphones and tablets, offset by a decrease in our embedded business. For GAAP, our outlook for the second quarter is as follows: revenue of $990 million to $1.05 billion, gross margin of 51.2% plus or minus 1 percentage point, OpEx of approximately $418 million. Our GAAP OpEx outlook for Q2 includes a onetime charge related to our corporate donation to Stanford Hospital of $25 million payable over a 10-year period. We approximate our tax rate for the quarter at 20% plus or minus 1 percentage point. For non-GAAP, we expect the following differences from our GAAP outlook: gross margin of 51.5% plus or minus 1 percentage point, OpEx of approximately $354 million, which excludes the $25 million onetime charge related to our contribution to Stanford, as well as stock-based compensation and certain other charges related to acquisition. Outlook for the second quarter depreciation and amortization are expected in the range of $55 million to $57 million and capital expenditures of $35 million to $45 million. Diluted shares are expected to be approximately 628 million.

Thanks, Karen. That concludes our prepared remarks. We'll now take questions. Ashley?

[Operator Instructions] And your first question comes from Ross Seymore with Deutsche Bank.

Just looking for the forward guidance, what are your segments doing specifically between the 3 main segments of your business?

Well, I guess they're all up.

Are there any specific drivers?

I'm sorry?

I thought you were done. Go ahead.

Yes, we're expecting them to be all up. Supply is still constrained, and our manufacturers are working their hearts out to catch up to our demand. But at the moment, demand still continues to exceed supply, and we expect that to happen for throughout the year as Chris was saying earlier. But Kepler is a fantastic, fantastic GPU. This is arguably the best we've ever made, and we've made some pretty fantastic GPUs over the years, and we're just in the process of rolling out Kepler into each and every one of our businesses and -- starting with desktop. There are still many, many segments within desktop that we're expecting to transition into Kepler throughout the year and throughout next year. There's a notebook that's in the process of rolling out now, and we started shipping this quarter and that -- or this last quarter, and that explains the big jump up and we're expecting another jump up this coming quarter. But Kepler is our GPU that goes into GeForce and Quadro and Tesla. And this upcoming week at GTC, I'd love to tell you about some of the new ideas around GPUs, and so Kepler is the core of many of our businesses. And when our GPU architecture is incredibly successful, you could see it drive all of our businesses for several years. And so my guess is that this will do again the same. And of course, Tegra 3 is ramping as we talked about before. Although we don't have LTE at the moment, LTE is predominantly a U.S.-centric phenomenon. Outside of the United States, the big superphone upgrade is quad-core. And Tegra 3 is the world's first quad-core and it's even a special quad-core at that with 4-PLUS-1, and it's got NVIDIA's branded graphics and all of our style of doing wonderful graphics and the video games and the content really shows. And so Tegra 3 is the next major upgrade for superphones in Europe and in China and elsewhere. And then towards the end of the year, we hope to bring Tegra 3 with LTE phones to United States.

I think this is my one follow-up. On that 28-nanometer shortage side of things, do you expect the solution to that problem to be more yield-driven by your primary supplier or -- I think in the past you've talked about looking to potentially diversify to additional foundries, so would the supply coming on from those potential suppliers fix the problem as well?

Yields of 28 are probably the best of any new node that TSMC has ever done. They've done a great job with 28, and I think that it also explains the demand on their 28. It's yielding wonderfully at this point in its ramp. The performance is fabulous. The efficiency, energy efficiency, is terrific. So I think it explains the reason why the demand is so great for this particular node. I think the increased supply will have to come from increased capacity and it's -- we're short now, and every chip that they can come out of their fab is being shipped instantaneously. We've reduced cycle time dramatically between us and our customers, and we're trying to get the products to market as quickly as possible. And so I expect us to be supply constrained from wafers at the wafer level, and we'll continue that way throughout the end of the year.

Your next question comes from Daniel Berenbaum with MKM Partners.

Maybe just to follow up on Ross' question a little bit. So your 28-nanometer yields are good and we've heard this from other -- from some of your peers as well who are also constrained by TSMC supply. Was the problem that the customers of Taiwan Semi just didn't give a bullish enough forecast so that TSMC didn't add sufficient capacity, or was there sort of a misunderstanding in the planning process? Can you maybe help us understand a little bit of the dynamic that's going on between you and your supplier?

Well, I think clearly, clearly, there just isn't enough capacity, and we've expected a really, really successful Kepler launch. And the reason for that is because we've known for some time that Kepler is going to be a fantastic GPU and we've had big plans for Kepler, and we invested last year rather aggressively on the various designs of Kepler and all of the capabilities that people were starting to learn about Kepler, and there are many more to come. And we made those investments because we expected to be able to grow the market, and there's every evidence that we're growing the market. There's just not enough capacity and the planning process when we think about these new nodes, it just simply needs to be better. And I don't think there's any way around that. In combination between us and TSMC, we under planned for the supply of 28-nanometer, and we need to fix that in the future.

So is this something that can simply be fixed by better planning, or is there really sort of a more structural change where TSMC is the only option at 28-nanometer, at least for some period of time, and getting the 20-nanometer will clearly have a lead among the other foundries of 20-nanometer. Is that a structural change where you're seeing potentially lower margins because you're much more tied to just the one supplier?

Well, I don't know that it's a structural change in the sense that if TSMC and ourselves were to have properly planned for this ramp, we wouldn't be shorting our customers so severely at the moment. And if we can make plenty of high-end GPUs like Keplers, our margins would be fine. And so it's really a shortage of supply that is causing our margins to be compressed at the moment, but we expect that to be cured as we go through the year. In terms of planning, that's a -- and building capacity, that's something that only TSMC can decide and ultimately, the industry will sort itself out.

Your next question comes from Harlan Sur with JPMorgan. Harlan Sur - JP Morgan Chase & Co, Research Division: Jen-Hsun, how much did the Tegra business grow in the first quarter? I know you were expecting it to grow at least 50%, and I'm assuming Tegra is going to contribute to the growth in the July quarter. Is the growth there going to be more driven by customers ramping new smartphones or new tablets or a combination of both? And then I have a follow-up.

Tegra grew 50% or more as we expected this last quarter, and we're expecting it to grow again this coming quarter. We launched the first phone in one region. The first phone was HTC's One X and the region was in Europe. Now it's expanding in Europe, and we're expecting to go through more regions with HTC. And then there are many other Tegra 3 phones that are teed up and coming to market. And so we're expecting the phone segment of Tegra 3 to continue to grow. There's also tablets. Tablets are doing very well for us. It represents about half of the overall demand at the moment, and there's some pretty exciting tablets that are about to come out throughout this year that we're working on. One particularly is, of course, Windows on ARM, and the enthusiasm behind it is growing every day. So it's a fantastic tablet, and I can't wait to show it you. And so I think the answer is to your question is yes to both. We're expecting to see growth in both tablets as well as phones. Harlan Sur - JP Morgan Chase & Co, Research Division: Great. And then obviously, you mentioned in your prepared remarks about under shipping demand in both the desktop and notebook segments in Q1. Can you just quantify roughly dollar-wise how much more revenue you could have shipped in Q1? And given obviously the constraints here in Q2, what the rough dollar impact on July sales could be as well.

We haven't broken it out, but it's a lot.

Your next question comes from C.J. Muse with Barclays.

I guess first question -- I guess trying to understand both the impact of the shortage of 28-nanometer but also you guys moving from good die to full wafer pricing, how should we think about the impact to gross margins today and literally, I guess, through the year?

Well, when we don't have enough of our premium products, then that will hurt our margins the most. And so I think at the moment, it's safe to say that supply is what's constraining, what's holding our margins back. As supply improves and yields will be improving throughout the year, I think we have every chance and every possibility of delivering great margins again. And so I think the #1 issue for us right now is just to make sure that we have enough supply of Kepler GPUs. And if that comes through, the margins will take care of itself.

That's helpful. And as my follow-up, can you talk about the attach rates with Ivy Bridge ramping? Usually, you talk about design wins. Would love to get some color there.

We expect to gain share in Ivy Bridge and we expect Ivy Bridge to likely be the best node, if you will, best processor platform that we've ever participated in. Primarily, the reason for that is because Kepler is such an energy-efficient GPU and the performance and the energy efficiency of the performance is just the best in the world, and I think that people are recognizing that. Ivy Bridge hasn't really changed anything from our perspective. The market structure seems to be consistent as it was before. Consumer demand continues to be quite high and consumers -- one of the most important applications for it is video games, and video games still prefer to have GeForce. And so I think all of those dynamics are still consistent. OEMs still would like to have ways to differentiate their products, and adding the GeForce processor has been one of the best ways to differentiate and create a premium segment and it remains that way. And so whether it's thinner notebooks or Ivy Bridge, I don't think that, that's really changed the dynamics. So we're expecting attach rates to be quite healthy and consistent with the past.

Your next question comes from David Wong with Wells Fargo.

When you talked about yields on 28-nanometers being so good at the moment, so as you do ramp Kepler and it matures somewhat, do you expect to get higher gross margins on this than prior-generation 40-nanometer chips? And also, are the shortages you're seeing on 28-nanometers translating into any upward pressure on wafer prices, or have you got pricing locked in for the Kepler generation?

The second question, I kind of lost a couple of bits. What did he say?

He said because supply is short, are wafer prices going to go up or are prices locked in.

I think the second question I have a better answer on. We have an understanding about wafer prices from TSMC and we negotiate with them and we now both understand where the pricing is, and so I don't expect a pricing increase and I surely hope it doesn't happen. With respect to margins, the reason why margins are great on Kepler is because it's a really efficient GPU. It's higher performance, it's lower power and for every performance level, it's also very efficient in terms of die size. And if that's the case, then our margins would lift. It is the most efficient architecture we've ever done and certainly relative to Fermi, our last-generation GPU, Kepler is much more efficient. And as you know, efficiency drives productivity, and in this particular case, productivity is margins. And so I expect that the more Kepler we ship, the higher our margins would lift. And so the really big focus for us is just to continue to ship as many Keplers as we can.

That's great. And also on Tegra 3, you said that it's roughly 50-50 smartphone and tablets. Given your current design wins as again as the year progresses, do you expect that to stay 50-50 or is it going to be more skewed towards either smartphones or tablets?

Windows on ARM is going to be the big variable on that. And at the moment, it's running 50-50 and about. And throughout the next quarter, my expectation is that phones would exceed tablets. But towards the end of the year, latter half of the year, the question is how does Windows on ARM do. And so that's the biggest lever, and we can report on that a little bit later.

Your next question comes from Brendan Furlong with Miller Tabak. Brendan Oliver Furlong - Miller Tabak + Co., LLC, Research Division: A follow-on, on the handset business. If you could, just maybe as the year progresses here, give us some indications on your baseband business, particularly on the LTE part of that.

Let's see. Well, we're really targeting the superphone segment of the marketplace. And when you look at the superphone segment of the marketplace, call it 100-plus million units a year or so, that part of the marketplace really is shared by 2 -- from a merchant semiconductor perspective, 2 players. Our focus -- because we didn't start from the modem business, we really focused on computing and graphics, we focused on segments of the marketplace where LTE isn't necessary at the moment and where it's just not ready for LTE. And so Europe, Asia, basically everywhere outside of United States has been our primary focus. Towards the end of this year starting in Q3-ish, we're working with LTE vendors and LTE partners to bring Tegra 3 and LTE phones to the rest of the markets. And so our primary focus at the moment is it's outside of United States. In the second half, in the United States, we'll bring LTE and Tegra devices into United States. Brendan Oliver Furlong - Miller Tabak + Co., LLC, Research Division: Excellent. And then my second question would be on the consumer side, if you could give any sort of indication of what the royalty piece -- portion of that is in the quarter.

You're talking about Sony there, Brendan? Brendan Oliver Furlong - Miller Tabak + Co., LLC, Research Division: Yes.

We don't disclose that.

Yes, we don't break it out.

Your next question comes from Romit Shah with Nomura.

Jen-Hsun, if you look at the PC guys that reported in April, they guided up a few percent, which was certainly better than seasonal. My question is on your Consumer GPU business, how much of the strength are you seeing is really driven by the channel restocking versus some of the share gains you've outlined with Kepler?

There's really 2 factors that's driving our growth, and one is 28-nanometer supply. There's just more of it than last quarter. I would say last quarter was not a typical quarter for us. I would characterize last quarter as a difficult quarter, and it was completely supply constrained. And this coming quarter, we're still supply constrained. And although it's a significant growth quarter, it is still difficult from my perspective, and the supply is still far short of demand. And the second part of our growth comes from Tegra 3's ramp, and we're really fortunate with Tegra 3 to have chosen an enhanced version of the 40-nanometer process. This is a technology called LPG, and we're the only one in the world using it. And it is a balance between high performance as well as extremely low power. TSMC did a fabulous job with us on that, and I really appreciate it because now we can support Tegra 3's ramp without consuming more of the 28-nanometer wafers that we can allocate to Kepler. And so I think the answer is, quite simply, more supply and more supply of 28-nanometer and Tegra 3's ramp that's contributing to the growth.

All right. My follow-up is on the cash balance. If you consider how the stock performed during the quarter and how much cash you generated on your balance sheet, if the cash balance is about 40% of your total market cap, and my question is are you managing or do you plan to manage the cash to a certain level considering the valuation? What are your thoughts around doing a share repurchase program or instituting a dividend?

We haven't -- we don't have anything to announce today, and that's a topic that we have with the board on a regular basis. But we don't have anything to announce today.

Just if I may, I'd like to give a little more clarity on the previous question from Brendan. On CPB, the royalties were flat and embedded, as you know, and our CFO commentary is down. So if directionally, that helps you.

Your next question comes from Ambrish Srivastava with BMO.

Jen-Hsun, I had actually 2 questions for you. The first one was we were all in a room all day yesterday at a small company in Santa Clara and there was a lot of talk about the structural advantage that Intel had going down in node. So just wanted your perspective since you do -- are one of the fabless companies that uses leading-edge technology and ultimately the first and the follow-up to that -- kind of related to that was as we go down in node, and some of the fabless companies have talked about going down in 20-nanometer, and what does that do to potentially margin or affects headwind. So what is your perspective on both those topics?

Well, I think our partner for advanced nodes is doing quite a nice job, I think, and I expect TSMC to continue to be able to advance their capabilities on an ongoing basis. There's nothing that would suggest that TSMC doesn't have the scale or the capabilities to do so. They're manufacturing an enormous number of wafers by anyone's standards, and so I think the availability of leading-edge wafers is not something I'm overly concerned about. In terms of process technology, it's also important to recognize that CPU's needs for process technology and SoC's needs for process technology is similar but not the same. And TSMC's leadership and density is something that can't be ignored. People always talk about one dimension of the semiconductor process, which is the transistor, but metallization is increasingly, especially for SoCs, by far, the most important variable. And so I think that we just need to be careful not to get -- drink too much marketing bathwater when we think about the technology variable because it's so complicated and SoCs are very different than CPUs. And so let's see, the second is margins. Our job is to make sure that we continue to create efficient architectures that are highly valued, and I think that if we continue to do and create GPUs like Kepler, I think we're going to be doing just fine.

Your next question comes from Doug Freedman with RBC Capital Markets.

Could you help us understand the dependency on 28-nanometer? What percent of your mix is it now? And at what point do you think we'll cross over to, say, the 50% of your output coming from 28? And I have a follow-up.

I think we're expecting it to be about 30% this coming quarter. We'd like it to be a lot more.

And can you give us any idea when you think you can reach 50%?

Well, this coming quarter's about 30% is my estimate. Next quarter will be more than that, and the quarter after that will be even more than that. And so hopefully, we cross 50% as soon as possible. We would like to cross this as quickly as possible.

And then my follow-up is really sort of an add-on to the previous question, and that relates to the efficiency that you're gaining at the advanced node. If they're so -- I guess I share your commentary that there isn't much risk of development of advanced process technology. I think the question comes into the economic advantage that the advanced technology is going to deliver. Are you getting the cost reduction? Clearly, 28 appears to be delivering that. Is the forecast for the next node going to give you an equal cost-to-benefit ratio or are you starting to see sort of a slowing of benefit as we go to these advanced nodes because they are so complex?

Every -- by the way, every node is complex. This is a conversation that we have at every node and we take every node seriously. And there's no evidence that anyone is becoming increasingly complacent, and so I think these are typical conversations to have. With respect to efficiency or productivity of each node, we will go into the node when it delivers value for the products that we need to create. And so if it makes no sense to go to a new node, we wait until it make sense to go to a new node. And so there isn't anything that -- nobody is in automatic cruise control, and we look at each node carefully with TSMC and we try to adjust each node if we can improve it for our products and otherwise. And then based on those analysis, we decide which products and when should go into the new nodes. And so it just simply needs to be analyzed. I don't -- there's nothing that suggests that the next node is going to be unproductive. If it's unproductive, it won't get used, and so my expectation is that it will be productive.

Your next question comes from Uche Orji with UBS.

Just real quick. As we look at the gross margin issues, can you talk about -- because one of the areas you've highlighted in the past is PSB and the future opportunities that, that brings. What are the gross margin drivers for going down the line, and I'm talking about looking at other areas, so if you can talk specifically on PSB. And also on Tegra, as these constraints happen across the group and then Qualcomm talked about it, do you see yourself increasingly gaining share going down the line on Tegra 3? That's my first question, please.

Okay. First, I'll answer the question backwards. Well, the answer is yes, we're gaining share. And the reason for that is because we were a very small player before. And as a discrete application processor provider, we've gained quite a bit of share. If you look at their market, our competitor's market -- business around discrete application processors, I think that you'll find that our share is growing quite rapidly. And so my expectation is that as we grow throughout the year that we're doing so because we're growing share, number one. Number two, PSB's gross margins are healthy and slightly improving, and the reason for that is because we continue to add more value to the high end of the visual computing market. Digital content creation is a growing market, and more and more of design is moving to digital forms, and this is an area that we are world leaders at and this is an area that we continue to add a lot of value to the industry on. And so as we increase value to the marketplace, we'll reward it with higher margins. And our focus is to make sure that PSB, where we have extremely high share already, that we continue to help grow the market. And that's one of the focus next week of GTC. If you all have a chance to either tune in or come even down to GTC, you're going to find some really, really fantastic work being done around the world by our partners and content developers who are using our technology to create wonderful assets and create wonderful things and discovery of new science.

Right. Let me just ask you a second question. Can you give us any updates on Windows on ARM as well the efforts in terms of timing of products coming, especially as Intel is making a big push in that category as well on the tablet front? So any commentary on development on the PC form factor on tablets and Windows 8 and regarding timing as well, that'd be helpful. That's one question. And then two, Kepler, it's been said that you will need Kepler in volume to be able to increase discrete attach rates in Ultrabooks. Any commentary as to the attach rates in Ultrabook in general? And how much of that is going to be Kepler? And how much of that will be a print [ph] products? That'll be helpful.

Windows 8 tablets and Windows RT tablets are very different things. Windows 8 x86 tablets and Windows on ARM tablets are very different things. And as Microsoft goes to market, people will understand why we're so enthusiastic about Windows on ARM. They're just very, very different things, and so I'll let Microsoft tell you about their plans, but I'm very enthusiastic about it. I think it makes a lot of sense for enterprise. Anybody who has a large part of their work around Windows would really benefit from Windows on ARM tablets and all the Windows on ARM devices -- range of devices that will be following after that. We've shipped about 1,000, maybe more than 1,000 Tegra Windows on ARM, or Microsoft has shipped more than 1,000 Tegra Windows on ARM tablets for software developers, and that number is growing very rapidly. And so we're working very closely with them to march to market with an exciting product as quickly as we can. Ultrabook, we see no difference between an Ultrabook and a PC. It's a thinner PC, but PCs have been getting thinner. And there's plenty of room from a thermal or mechanical or electrical perspective to support a GPU. And the reason why people put GPUs in their notebooks is because it helps them deliver a premium device. It also enables consumers the ability to enjoy multimedia better. And so I don't see any difference relative to attach rates when it comes to Ultrabooks. And then with respect to Kepler, Kepler is our primary focus for next-generation notebooks and we're trying to build as much as we can. And the more that our supply catches up to the demand, the more our customers can ship Kepler GPUs. Everybody is really, really anxious to get more, and so we need to just make more.

Your next question comes from Daniel Gelbtuch with Cantor. Daniel M. Gelbtuch - Cantor Fitzgerald & Co., Research Division: Just digging a little deeper into Tegra. I was just wondering if you could compare and contrast how Tegra really, not only from a efficiency perspective, how it compares with the current status quo in the market, number one. And number two, with regard to SDR or software-defined radios and Icera, how do you think that changes the landscape for LTE, I guess, next year and the year -- and beyond?

Well, Tegra 3 is the world's first quad-core processor. And it's not only quad-core, as you might know. It has a really energy-efficient CPU core as well. And so not only does it deliver higher performance, it's also more energy efficient than the status quo. And of course, with all of the products, the graphics and the video games are just so much better. And so I think that based on the demand and the reception in the marketplace for the first Tegra 3 phone, I think it's safe to say that Tegra 3 is really the standard by which everything is being compared, number one. Number two, software-defined radio, there's no question that SDR allows us to deliver a much smaller and efficient and cost-efficient modem. And with LTE, it is very clear that the die size is smaller, the energy efficiency is excellent and that's really going to help, especially with capacity being so constrained. Having a very small modem, a very small LTE modem, is going to be a big differentiator for us. Not only will we be able to make more of integrated versions of application processors and LTE modems, we'll also be able to bring more cost-effective solutions to the marketplace. So we're really, really gunning for that and running as hard as we can. Daniel M. Gelbtuch - Cantor Fitzgerald & Co., Research Division: So is it fair to say that once you get this integration of SDR rolling, this would be perhaps the first time that the current incumbent in the APE space is going to be seeing some serious -- a real competitor?

Well, the incumbent is a really good company, and they're very competitive. And this is new for us, and we still have a lot to learn and we're learning as fast as we can. But the Icera SDR modem is really revolutionary and when we can combine that -- and there's every evidence it works great. We're shipping phones now. There were people said that we couldn't get that modem to market, people said that we couldn't get phone's voice to work. But obviously, all wrong. We're now certified on LTE and we're in the process of getting products to market. And so these are all barriers that we have to overcome, but we're taking it -- ticking it off one at a time and we're -- I'm cautious, but very enthusiastic about the technology we're bringing to the market.

Your next question comes from Suji De Silva with ThinkEquity.

So first question, Jen-Hsun, on this Kepler architecture and the energy efficiency of it as you described it, how do you monetize that? Is that smaller die or is that a value proposition that customers pay for and just allows you to kind of get that upgrade?

Well, absolutely both, absolutely both. Kepler is the most energy-efficient. It's also the most die size-efficient GPU we've ever made. The performance per watt and the performance per millimeter are both the best that we've ever done. It was a huge drive for our company to transform ourselves in just about every single dimension from architecture to design to implementation to software. We pulled everything together for Kepler. When you're more energy-efficient, at any level of power that the system can provide will be higher performance. And so if the system can provide 300 watts, we're higher performance. If the system can provide for only 10 watts, we're higher performance. If the system can provide 1 watt in the future, we'll be higher performance. So energy efficiency is incredibly important. And of course, at any price point in the marketplace, if we are efficient with respect to die size, that can only help our gross margins. Whatever gross margins we had at the Fermi generation, so long as market dynamics remain similar, I expect our gross margins to be higher than the Fermi generation wherever that happens to be. And so that explains why I'm so impatient to get Kepler lined [ph] up and why we need so much more Kepler supply from TSMC.

Understood. And my other question, Jen-Hsun, is in January, you had a Microsoft representative in CES and you asked them about how Windows -- Microsoft would avoid confusion as Windows on ARM came out, and you may defer this question to Microsoft, but what's your update on that thought as we look ahead to Microsoft tablets?

Well, I think you guys are going to really like the answer. They've been very thoughtful about the segmentation of the Windows -- the next Windows generation. There are different customers with different needs, and I think you're going to find their answer rather thoughtful. I'm very pleased with where they ended up and I think it's going to really, really, really help the positioning of the Windows on ARM tablets because they're very different than those -- than x86. It's a new product category. I think you're going to find that they're really desirable, and they're going to bring a lot of value to the marketplace. For people who -- where Windows and Office is still a very central part of their digital experience, I think you're going to find that this is really the best tablet on the planet.

Your next question comes from Raji Gill with Needham & Company. Rajvindra S. Gill - Needham & Company, LLC, Research Division: The question I have is on Tegra. I just wondered if you can update us on the annual guidance that you had given at the start of the year. Any changes to that, any additional color?

Thank you. No, no change to that. No additional color. The 3 growth drivers for Tegra are: One, phone design wins; and two, because we've successfully reduced the costs of tablets, we're starting to reach price points of Android tablets that are really resonating with the market and of course, the Android tablets are just getting better and better all the time, and Tegra 3 Android tablets are really delightful. And there's a whole bunch of new tablet that are coming starting now and towards the end of the year. And then the third driver is Windows on ARM. That's coming together really nicely, and I think that people are going to be quite delighted by it. And so those are really the 3 primary drivers and they, all 3, remain drivers at the moment. Rajvindra S. Gill - Needham & Company, LLC, Research Division: Just a question on the OpEx. The OpEx went up about 7% sequentially. It's being guided up again. So now we're kind of at the high end of your OpEx range. How should we look at the OpEx throughout this year? Do you think you could really go past that range? What -- how do you think we should look at it next year? And along those same lines, are we starting to be kind of a return on investment on the Icera?

We are going to do the best we can to stay at the guidance level that we provided previously. Now if there -- if some of the new initiatives in our company starts to gain traction, we'll increase the level of investment to take that to market. But we'll play that as we go. But at the moment, our current plan is to stay as close as we can to the guidance that we provided before.

Your next question comes from Craig Berger with FBR Capital Markets. Craig Berger - FBR Capital Markets & Co., Research Division: Can you help us understand the impact Kepler's having on GPU ASPs and what ASPs did in desktop and notebook in the first quarter? And then I have a follow-up.

Well, the more -- because Kepler comes in through the performance segment and the enthusiast segment, it naturally lifts ASPs. The more we ship, the more -- the higher our ASPs will be. And so Kepler ASPs are higher than Fermi ASPs because they target the premium segments. Kepler margins are better than Fermi margins because they're in the enthusiast segment and the performance segments, also because they're so much more energy-efficient and so much more die size-efficient. And so I think the simple answer is we just need a lot more Keplers, and the demand for Kepler is really, really high in desktops, in notebooks. The gamers come out and refresh their PCs once every few years. And on all accounts, Kepler has been a complete home run. And this is -- it's pulling people out to refresh their PCs, it's -- and PC gaming is really vibrant right now. And so I think those factors are really driving Kepler. So the more we can build, the higher ASPs we will be and surely the much higher margins we will be. Craig Berger - FBR Capital Markets & Co., Research Division: As the follow-up, can you update us on your CFO search? And then really I wanted to ask again on the OpEx. I know it was just asked, but can you just reiterate what was your guidance? And you said you might spend more; is that more than you're spending now, more than your guidance? And what's the time line there because, I mean, from my perspective, you did better on revenues, you did better on gross margins and my model is still challenged because of the OpEx. And so that's kind of disappointing.

The guidance that we gave you on OpEx is exactly the guidance we gave you on OpEx. And so we just provided the guidance and that's what it was, and that's what we're planning. With respect to -- what was the first question, CFO search? I think my CFO is doing fabulously, and she's doing a fantastic job with the business. And we're looking. And if there's a world-class CFO that we fall in love with, certainly we'll consider. But Karen's doing a fabulous job.

At this time, there are no further questions.

All right. Thanks, Ashley. Thanks, everyone, for listening in, and we look forward to talking to you soon about our second quarter results.

That concludes today's conference. Thank you for your participation. You may now disconnect.