NVIDIA Corp (NVDA) — Q3 2016 Earnings Call Transcript

Thank you. Good afternoon, everyone, and welcome to NVIDIA's conference call for the third quarter of fiscal 2016. With me on the call today from NVIDIA are Jen-Hsun Huang, President and Chief Executive Officer, and Colette Kress, Executive Vice President and Chief Financial Officer. I'd like to remind you that today's call is being webcast live on NVIDIA's investor relations website. It is also being recorded. You can hear a replay by telephone until the 12 of November 2015. The webcast will be available for replay up until next quarter's conference call to discuss Q4 financial results. The content of today's call is NVIDIA's property. It cannot be reproduced or transcribed without our prior written consent. During the course of this call, 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 most recent Forms 10-K and 10-Q and the reports that we may file on Form 8-K with the Securities and Exchange Commission. All our statements are made as of today, the 5 of November, 2015, based on information currently available to us. Except as required by law, we assume no obligation to update any such statements. During this call, we will discuss non-GAAP financial measures. You can find a reconciliation of these non-GAAP financial measures to GAAP financial measures in our CFO commentary which is posted on our website. With that, let me turn the call over to Colette.

Thanks, Arnab. Third quarter revenue was a record $1.305 billion, up 7% from a year earlier, up 13% sequentially and substantially above our outlook of $1.18 billion. Growth was driven by record revenue for gaming, particularly for our GTX platform and automotive infotainment systems. From a reporting segment perspective, GPU revenue was $1.11 billion, up 12% from a year earlier. Tegra processor revenue was $129 million, down 23% from a year earlier. Our strategy remains focused on creating platforms for gaming, professional visualization, data center, and automotive. Our progress is the result of product innovation, our strong position in growing markets, and disciplined execution. In Q3, our four market platforms contributed 85% of our revenue, up from 71% a year earlier. First, let's start with our gaming platform. Gaming revenue rose 44% year-on-year to $761 million. Growth is being fueled by a number of factors, among them, the significant increase in graphics production value, the rise of eSports, the anticipation of blockbuster games for the holiday season, and the emergence of new technologies like VR for richer, more immersive game play. Professional gaming competitions or eSports are well on their way to becoming a global entertainment category with an audience of 188 million viewers. There were 30 million viewers alone for last month's League of Legends World Championship. The launch of great gaming titles drives gaming platform sales. We are excited by the strong pipeline of AAA games ready for the holiday season, including Star Wars Battlefront, Call of Duty, Black Ops III, Rainbow Six, and Fallout Four. Meanwhile, we are looking forward to the launch of virtual reality, with the approaching availability of VR headsets from Oculus, HTC, and others, as well as new VR games such as EVE: Valkyrie. During the quarter we enabled a new category of high capable enthusiast class gaming notebooks with the launch of GTX 980. For the first time notebooks are able to deliver the same gaming experience as desktops and have the processing power to drive frame rates needed for VR. Our long-term bet on Android gaming is SHIELD. During the quarter, we extended the reach of SHIELD Android TV, the market's most advanced 4K smart TV console in two key European markets. We launched GeForce NOW which provides Netflix-like service for streaming games from the cloud to your SHIELD device. Moving to professional visualization. Quadro revenue increased 8% sequentially to $190 million, a decline of 8% year-over-year. At the SIGGRAPH Professional computer design conference, we introduced NVIDIA DesignWorks. This suite of algorithm libraries and tools enables developers to easily incorporate photorealistic visualization of product and building designs. And with DesignWorks VR, developers can easily provide the amazing experience for such areas as theme park design, product design, telepresence, and surgical training. In data center, inclusive of Tesla and GRID, revenue rose 13% sequentially to $82 million, reflecting demand in deep learning, high-performance computing, and graphic virtualization. Revenue declined 8% year-on-year due to the lumpiness from large deep learning projects. Most of the world's leading web services companies have adopted Tesla for deep learning to power next-generation services. In addition to using our GPUs for training servers on massive data sets, companies like Baidu are using them for inference for such work as real-time voice recognition. In high-performance computing, we increasingly find that GPUs are moving from the domain of researchers to commercial use, in areas such as weather forecasting and seismic processing. We continue to make progress as well with our GRID virtualization platform, which enables companies to deliver graphic-rich applications to employees on any device anywhere. At VMworld in San Francisco, we launched GRID 2.0, with twice the user density and application performance of its predecessor and support for both Linux and blade servers. A dozen Fortune 100 companies are completing trials of GRID 2.0 and more than 125 OEM offered servers are qualified to run it. Also reflecting momentum for NVIDIA data center offerings, Microsoft Azure announced that it has integrated NVIDIA GRID into its rapidly growing cloud service in response to customer demand for GPU acceleration in the Cloud. Finally, automotive revenue rose 51% year-on-year to a record $79 million. At last month's International Auto Show in Frankfurt, Mercedes Benz, Audi, Porsche, Bentley, and Honda all showcased a range of production vehicles and concept cars with NVIDIA-powered digital cockpits, and our partner, Tesla Motors just introduced its falcon-winged Model X. Like the Model S, the Model X is equipped with both a multi-touch infotainment system and a digital instrument cluster powered by NVIDIA. Finally, we are continuing deeper collaborations with over 50 companies that are developing self-driving car technologies using NVIDIA DRIVE PX. These include car manufacturers, Tier 1 OEMs, start-ups, and research institutions. DRIVE PX is an advanced autonomous driving platform that fuses data from cameras, lidars, radar, and ultrasonic sensors to determine the path forward using deep learning techniques. We look forward to sharing more details about this work at CES 2016. Our OEMs and IP business was $192 million, up 11% sequentially, driven by seasonal demand for notebooks. Now, turning to the rest of the income statement. GAAP gross margin was 56.3%, non-GAAP gross margin was 56.5% in line with our outlook. GAAP and non-GAAP gross margin increased from a year ago reflecting our strong gaming revenue. GAAP gross margin increased sequentially, as the previous quarter included a charge related to the NVIDIA SHIELD tablet recall. Non-GAAP gross margin was nearly flat sequentially. GAAP operating expenses for the third quarter were $489 million, inclusive of $8 million of restructuring and other charges. Non-GAAP operating expenses including litigation charges were $430 million, slightly lower than our outlook. We continue to focus on operating efficiencies, and we manage investment growth while growing operating margins. GAAP operating income for the third quarter was $245 million, up 15% from $213 million a year earlier, reflecting strong revenue growth. Non-GAAP operating income was $308 million, up 17% from $264 million a year earlier. GAAP net income was $246 million. GAAP earnings per diluted share were $0.44, including $0.01 of restructuring and other charges, as well as a benefit of $0.09 from a tax reserve relief. Non-GAAP net income was $255 million. Non-GAAP earnings per diluted share were $0.46, an increase of 18% year-over-year. Now turning to some key balance sheet items. In Q3, our cash and marketable securities balance was $4.73 billion. During the third quarter, we paid $53 million in cash dividends and we closed our accelerated repurchasing agreement with an additional 4.6 million shares returned. As a result, we have returned to shareholders an aggregate of $604 million fiscal year-to-date of our full year intent to return $800 million. As part of our ongoing commitment to deliver shareholder value, through capital return, we have announced an 18% increase to our quarterly cash dividend to $0.115 per share to be paid on December 14 to all shareholders of record on November 20. In addition, we are pleased to announce our intention in fiscal 2017 to return approximately $1 billion to shareholders through ongoing quarterly cash dividends and share repurchases. Now turning to the outlook for the fourth quarter of fiscal 2016. We remain excited about our business prospects. Gaming continues to accelerate, and eSports, VR, and a pipeline of exciting games will lift it further. GPU accelerated data centers are expanding in both high-performance computing and cloud, driven by the acceptance of deep learning techniques. And the transition to autonomous driving is well underway. We have excellent positions in each of these growth markets. We expect revenue for the fourth quarter of 2016 to be $1.3 billion, plus or minus 2%. Our GAAP and non-GAAP gross margins are expected to be 56.7% and 57%, respectively, plus or minus 50 basis points. GAAP operating expenses are expected to be approximately $503 million. Non-GAAP operating expenses are expected to be approximately $445 million. For fiscal 2016, we expect non-GAAP operating expenses excluding litigation costs to be approximately flat with fiscal 2015. Litigation costs are anticipated to be in the range of $70 million to $80 million as we defend our intellectual property. GAAP and non-GAAP tax rates for the fourth quarter of fiscal 2016 are both expected to be 20%, plus or minus 1%. The above GAAP outlook excludes restructuring charges which are expected to be in the range of $25 million to $35 million as we finalize the wind down of the Icera operations in the fourth quarter of fiscal 2016. Further, financial details including the CFO commentary and revenue by market platforms are available on our IR website. With that, we will now open the call for questions. Operator, will you please poll for questions?

Thanks for the question, and congratulations on the great results. For my first question, the last two quarters you have substantially exceeded your original outlook and PC gaming has been an important part of that. I'm wondering these trends that you're seeing or the upside surprise that you have seen in both these quarters, has that come more from units, has that come more from better mix or ASPs or share gains? And more importantly, how sustainable is this kind of upside trajectory?

Yeah. Vivek, first of all, thanks. Well, I think that we've been consistent that our PC gaming business is driven by several factors. The first factor is the great AAA titles that are coming out with ever increasingly graphics production value. The second is just the number of eSports players that are growing around the world. It’s a very social phenomenon. The more of your friends joining eSports, the more you have to join eSports, so you can hang out with your friends. The third is the developing countries. One of the things about PC gaming and overall gaming is that you really need to have broadband access anymore so that you can download these amazing games. As broadband adoption continues to grow in the developing countries, the PC is really the best way to get into gaming. It's a platform that you need for your daily lives anyways, and surely most young people need it for school. And so that's been helpful. All of those factors have increased both the units as well as the ASPs that we're enjoying. If you look at the size of the overall market, I think it's probably fair to say that we're not fully penetrated. Our expectation is that as we look into next year, all of the great games that are coming out, whether it's Call of Duty, Star Wars, or Assassin's Creed or Tomb Raider, surely everybody's quite excited about VR coming out in the beginning of next year. I think there are several growth drivers that are still quite powerful, and our expectation is that this is going to be a large market going forward.

Great. Thanks, Jen-Hsun. And for my follow-up question, not to quibble, but when I look at your Q4 guidance, it's strong, better than expectations, but it implies flattish trends in what is usually a seasonally stronger quarter. If you could just give us some color on what segments directionally do you think would be up or down versus Q3? Thank you.

Well, the guidance that we provide is the guidance that we think is best and balanced. We'll see how the quarter plays out. But the growth drivers in our business overall, gaming is a growth driver, and we just mentioned some of the dynamics that are underlying it. Data centers is a growth driver for us, and I'm sure we'll be talking more about that during the conference call. And then third, the automotive business is a growth driver for us. So we have several growth drivers, but at this point, our guidance is the best we have to offer.

Okay. Thank you, Jen-Hsun.

Hi. Thanks for taking my questions. First question on virtual reality as a driver. It seems that you have fairly low expectations for 2016. When you see the Oculus VR demos, I really leave with the sense that they could transform business processes for many different industries, and it seems like a lot of potential. So I'm just trying to reconcile that kind of reaction that I've noticed versus your, it seems like lower key expectations for virtual reality as a driver. So if you could help reconcile that, that'd be great.

Well, first of all, Mark, we are over the top excited about virtual reality. There's nobody at NVIDIA who is low key about it. There's no question that when you enjoy the Oculus headset with GeForce GTX 980 or GeForce GTX Titan the experience is unbelievable. Recently we even produced a whole family of notebooks that are able to enjoy VR directly from the notebook. We're super excited about VR. The sense of presence that you have in enjoying VR headsets whether it's from Oculus or Valve's Vive, it's really fantastic. We also see that the application of VR is quite sweeping. Of course, we're quite enthusiastic about games and the games that we've seen are really amazing. But it's used for industrial design, architectural design, it's used for medical imaging, even scientific computing. We're seeing entertainment, large-scale entertainment, large format entertainment. We’re tracking over 250 companies now that are working with us on VR and it ranges everywhere from video games to entertainment to professional graphics, as I've mentioned. Our approach is to provide a platform by which the entire VR industry can be built upon. We call it our DesignWorks VR for our enterprise products, our professional products, and GameWorks VR for our consumer products. By connecting to our APIs and connecting to our SDK, the performance and the experience is just phenomenal. Maybe your question is much more related to financial. Our approach there is to be enthusiastic about the work that we are doing, to support the entire industry as it moves into VR. We are engaged deeply with the industry in doing so. I think that we built the premier platform for VR for both PC gaming as well as professional graphics. As far as financial is concerned, realize that the shipment hasn't really started yet and I think it's prudent to wait and see. But my expectation long-term is that VR is going to be a very powerful growth driver for us. That is almost undeniable at this point based on everything that we've seen and all the effort that's going into it.

Thanks. That's very helpful color, Jen-Hsun. Thank you for that. Follow-up question, if I may. On the gaming software, the PC gaming software, there's a strong product cycle you've mentioned a number of times. Is there a risk that this is partially helping to drive gaming GPU sales now and then there's some kind of a let-up later on in, say, two or three quarters from now? How should we think about that? Is there a correlation between gaming software launches and lumpy shipments of gaming GPUs? Thank you.

I guess in the end, it's hard to predict. There are only a few things that we do know. The first principles of what we know are that the gaming market overall is growing. The number of eSports players is continuing to grow. The number of countries, developing countries, that are now able to bring more people into gaming is growing as they continue to adopt broadband. The production value of the video games are just amazing. Boy, if you just look at Star Wars and Assassin's Creed and Call of Duty and Tomb Raider, I mean, they're almost cinematic. Our expectation is that these factors combined are going to continue to drive the gaming market forward. We'll see from quarter-to-quarter how they do, but at the moment everywhere in the world, we're just seeing gaming to be quite vibrant.

Thank you.

Great. Thanks for taking my questions. The first one, Jen-Hsun, if I could on the automotive market. First of all, at your last Analyst Day you guys mentioned an aggregate backlog number for an estimate on the rough value of the design wins over a number of years; I was wondering if you could potentially provide us an update on how that number changed over the last several quarters. Related to that, given the nice pipeline that you guys have building up, can you talk about what sense of urgency your customers are telling you guys in terms of how quickly they need to get Tegra into some of these systems, whether it's infotainment or ADAS-type products, and what that translates to in terms of the revenues coming on for Tegra?

Yeah, thanks a lot, Stephen. First of all, our automotive business is growing quite robustly and year-over-year it's grown over 50%. Our expectation is that it's going to continue to grow through next year and a couple of years after that. Meanwhile, there are two very important dynamics that are happening in the car industry that are very favorable for us. One dynamic has to do with the central car computer. It is very clear now that your future car will be software-defined. There will likely be multiple operating systems. These operating systems will control graphics-rich experiences. This centralized computer will be built like a data center so that it can be secure, it can be robust, it can be mission-critical. Software is going to play just a huge role in it. So as the car becomes more computerized, more than just digital, more than just electronic, it becomes computerized and that it's software-rich, it plays into our hands because we're such a great software company. We know system software incredibly well, and software capability is one of our major differentiating advantages. It's one of the reasons why we're able to do what we do. The second dynamic is that whereas the first generation of driver assistance was made possible with radars, the second generation included cameras. The first generation of autonomous driving is about to come. Autonomous driving hasn't really made its way into the market yet. We're seeing the early versions of that, but architectures that enable fully autonomous driving are yet to come. This is an area where we really have a great deal of capability. It's going to require sensor fusion, it’s going to require an enormous amount of software, it's going to require artificial intelligence, and it's going to require the ability to support an open computing platform that NVIDIA is built upon. If you look at our accelerated computing platform, it's programmed by so many people in the world and this is a real advantage because car companies need the ability to develop their own artificial intelligence network so they can ultimately own their driving experience. The new era of autonomous cars and autonomous vehicles is a great opportunity for us. So those two dynamics are going to be fundamental to the future growth of our automotive business.

Okay. Thanks for that color, Jen-Hsun. And as my follow-up question, I was wondering if you could help parse out some of the drivers within the Tesla and GRID product families. Just between high-performance computing, cloud data center, and traditional corporate enterprise-type customers that are currently buying your Tesla and GRID products. Can you talk about the near-medium term demand trends across the different verticals? Thanks.

Yeah, I appreciate that question. PC gaming, of course, is a powerful growth driver for us. Our automotive business is a growth driver for us. I fully expect our data center business to be a growth driver for us. We've simplified our data center strategy over the course of the last several quarters. We used to have multiple data center strategies; we now have basically one data center platform with multiple applications that drive it on top. The data center platform, the accelerated data center platform is called Tesla. On top of it, we have multiple software stacks. The first software stack we created was for high-performance computing for scientists to be able to do things like weather simulations and molecular dynamics simulations, quantum chemistry simulations, and material simulations. HPC, high-performance computing was our first stack. We also talked about GRID. GRID is our enterprise virtualization stack. We have GeForce NOW, which is our cloud gaming stack. One of the things that’s super important in the future is the work that we’re going to do in hyperscale data centers that allowed cloud data centers to do their work around artificial intelligence and take the pressure off of the enormous amount of new traffic that is swamping and flooding data centers from users generating their own video. Everybody's going to become a broadcaster. There are PC gamers that are already broadcasting on a daily basis, and Twitch is the name by which we all know it. You're going to see YouTube Live. YouTube Live is doing incredibly well. You're going to see applications like Meerkat and Periscope and there are more coming that basically lets almost everybody become a broadcaster. All of those videos today need to be somehow processed. Our GPUs are really fantastic at that. We have multiple applications, if you will, that sit on top of our data center platform. The dynamics that I've mentioned are among the most important dynamics around data centers—Moore's Law needs a boost. We know now that Moore's Law is under some amount of pressure; it's not growing as fast as the industry needs it to. So Moore's Law needs a boost and that's really what Tesla is about. Tesla is an accelerated computing platform that boosts the microprocessor. So that's one dynamic. The second dynamic is what I mentioned; the global race to AI. There's not one data center on the planet. There's not one cloud service on the planet. Frankly, there's not one application company on the planet that we know of, that's not trying to figure out how AI can completely revolutionize the applications that they provide. Accelerating the cloud and making it possible for cloud to offer smart services and graphics-rich services is exactly what we announced recently with Microsoft Azure. Azure now provides the NVIDIA GRID to power graphics-accelerated applications as a service and we're super excited about that. So those dynamics are happening all in real-time and it's contributed to our sequential growth of 13%.

Okay. Thanks, Jen-Hsun and congrats on the strong results.

Thank you. Congrats, guys. Hey, Jen-Hsun, can you give us a quick rundown on the competitive dynamic in the data center, particularly with GRID? If there is any kind of competition and the reason I bring it up is that the Intel Altera combo perhaps is a way to kind of fight what you guys are doing there? Thanks.

Yeah, Hans, first of all, thank you. Well, I think it's very clear at this point that boosting Moore's Law is an essential path forward that relying on CPUs alone for the entire workload of a data center is unrealistic and frankly unnecessary. There are no supercomputing centers that I know of that are not considering accelerated computing. Relying on traditional methods unfortunately doesn't allow you to scale your data center efficiently anymore. There are no hyperscale data centers that I know of that aren't somehow overwhelmed and flooded by user-generated video content or the need to move to machine learning and artificial intelligence to provide smarter services. These dynamics are putting a lot of pressure on data centers and it's one of the reasons why there's so much talk about new approaches for improving the throughput of these data centers, whether it’s supercomputing data centers or hyperscale data centers across the board. This is an enormous opportunity for us. This is an area that we have invested almost 10 years in. I've been talking about accelerated data centers now for well, 10 years. We're seeing a lot of momentum and as other people pursue other ideas, I think it's just an acknowledgment that this is a good path forward. As for your question specifically about FPGAs, we use FPGAs at NVIDIA. We use FPGAs to do simulations and emulations of our products. FPGAs have been around for a long time and they are perfect for certain applications. The challenge for— I guess the tradeoff is an FPGA is like, it's designed. It's designed and configured on an array of transistors and gates, whereas a GPU is a full custom processor designed completely by hand. Whereas one FPGA is designed probably by a handful of engineers, a GPU is designed by thousands of engineers. A GPU performing a task could be 100 times the performance of an FPGA. It stands to reason that it's a result of handcraftsmanship. An FPGA, however, can do specific tasks quite well. If you only want to do one or two things, you can limit the functionality and make an FPGA quite efficient. However, our perspective is that a data center doesn't do one or two things; a data center runs thousands of applications. A data center has video, has imaging, it has voice, it does machine learning. Algorithms are changing every day. Today, the algorithms are convolution neural nets, tomorrow, it has memory, and after that, it's recursive. The type of neural nets just keep growing and the algorithms keep moving forward. Our belief is that the vast array of applications and the evolution of the algorithms really makes a processor much easier to adopt and much faster to increase the overall throughput of your data center. We'll see how it plays out, but at the moment we are enjoying a lot of enthusiasm around accelerated computing.

Okay. Thanks for the rundown.

Thanks for taking my question. I'm just wondering on GeForce NOW, how is that revenue going to be recorded as you start getting subscribers and is that going to go under GRID or is that going to go under SHIELD, and can you talk a little bit more about what gross margins are expected for that?

Well, it's not very much right now, so it's not much to talk about. But I think that the way to think about it is this, SHIELD and GeForce NOW is our long-term bet on the future of gaming. When we look out in the future, I think there are two dynamics that I think are unquestionable. One dynamic is that Android is going to become more important in an operating system. Don't forget gaming is about computer gaming. Gaming is made possible because it’s built on top of a computer architecture and Android is the most popular operating system in the world. It's undeniable that it will become a major force in gaming someday and SHIELD is our bet in that. The second is that cloud services application as a service, and therefore video games as an application, as a service will eventually come. It took many years for Netflix to eventually become mainstream, and my expectation is that it will take several years for GeForce NOW to become popular worldwide. Those two dynamics, I think, are largely unquestionable. The question is not so much if; the question is simply when. Then maybe the next question is who? Who is best equipped to be able to bring this future into the world and we feel that as the world leader in visual computing and as one of the major platform leaders in video games today, this is an area that we can really make a great contribution. It's a modest investment for us, and we are thoughtful about the level that we invest to ensure that it aligns with long-term opportunities. If we were to recognize the revenues, to put it back into the original question, if we were to recognize the revenues, it would go into gaming.

Okay, great, and maybe if I could ask one other question about the split in the PC gaming business between notebook and I've seen a lot of reviews of the new gaming notebooks; it seems all top 10 are based on GeForce. Between notebooks and your cards, what's the revenue split now?

Well, the revenue split is largely still PC desktops. I appreciate you bringing up the notebook work that we did. Maxwell, the GPU architecture that we created, and the craftsmanship of the GPUs we made are so incredible that it’s finally possible for a notebook to deliver the same level of performance as a desktop. Our timing was timed so that people who want to enjoy VR with a notebook can finally do it. The latest generation of notebooks with GTX 980 are just amazing. They are so many times more powerful than a game console, fits in a space smaller than a game console, and can drive VR. Everything you want to do, every game you want to play, is possible on that thin laptop. The enthusiasm behind our launch with GTX 980 has been really fantastic. I appreciate you recognizing it.

Well, congratulations.

Hi. Thank you. Two questions. One is back on the professional side, Jen-Hsun. If I look at Quadro and Tesla, what reverses the year-over-year decline that we have seen the last couple of quarters? As it relates to Tesla, is it just a matter of timing of when the HPC systems get launched? The second follow-up was for Euclid on the IP side and expenses for IP. It's not debatable that you guys have a very unique IP portfolio, but I'm wondering given the initial reversal you had on the lawsuit against Qualcomm and Samsung, why should the expenses be continuing to be higher on that front? Thank you.

Sure. Thanks for the questions, Ambrish. Let me take it in reverse order. So first on IP. It's not over. The administrative law judge has opined that of the three patents that he had to rule on, that two were not infringed and one was infringed but he opined invalid. Of course, we disagree with that and we have submitted our rebuttal. All of that has to go to the commission, which is ultimately where the ruling happens. The next phase is for the commission to decide whether they want to review, and we hope they do. It's not uncommon for them to do. In February timeframe, the commission would determine their judgment, and it is not uncommon at all to have the commission disagree with the administrative law judge. I think it's far from over and therefore it's not over. But I think the higher level point is this: Our IP is very important to us. That invention as the world leader in our field is fundamental and core to our future. We have to invent the future. We are the world leader, after all. From time to time, we may decide that as a one-off and strategically decide that we would defend our IP. And this is what we have done in this case. But it's not our business model. We don't litigate for our business model. We don't depend on licensing for our business model. I hope you all on the call today don't invest in our company as a result of IP royalties. What I hope you see is that we are quite a unique company, and we are a world leader in the space that we endeavor. Visual computing is more important than ever and applies to more industries than ever. Our business model is really about selling products into vertical markets that are really exciting and quite vibrant and growing. I hope that you decide to invest on the basis of the products, the business model, and our special position in the marketplace. On the two questions related to Tesla and Quadro, as you know, Quadro is about design and it's about workstations, it's about manufacturing, it's about things that people make. We are a very, very, very significant leader in this market. We make a great deal of contributions to this market, but it's also a mature market. Our strategy for growing Quadro is very simple. We have to reinvent it, and we have to reinvent it by bringing new capabilities to it, and we have two that we’re quite excited about. One is a new technology called Iray, and Iray is a brand new way of generating images that is photorealistic and physically accurate. It's almost like inventing the font of desktop publishing. What you see is what you will get. It's just really quite amazing. I think Iray is going to rejuvenate the way that people do computer graphics and I hope it increases the size of the market, makes it easier for people to design products and increase our ASPs quite significantly. Number two, VR. We have mentioned before that from design to architectural design to medical imaging to scientific computing to large-scale entertainment, we're seeing that VR is going to be very important. Those two we hope to drive the future growth of Quadro and get it back on a growth track. As for Tesla, we grew 13% quarter-over-quarter. We declined year-over-year. Frankly, it didn't meet my expectation either. However, the fundamental dynamics, the reasons don't really matter here or there, but the fundamental dynamics are incredibly sound. The groundswell around the fundamental dynamics is unquestionable. Moore's Law needs a boost. AI, every company on the planet is racing towards the future of AI and cloud computing needs to be accelerated with graphics as Microsoft Azure has demonstrated. My expectation is that we'll grow sequentially again and that the fundamental dynamics around Tesla are really, really positive.

Thanks for all the details, Jen-Hsun. Good luck.

Yeah. Thanks a lot for taking my question. On virtual reality, so the recommended specs from Oculus are something like a GTX 970, but I think a lot of third parties and I think even you guys have mentioned that it actually requires much more in order to get a pretty reasonable experience. So can you help us understand what we should be expecting in terms of the average set-up for what's called an early adopter of Oculus? Should we expect two GTX 970s, should we expect a GTX 980, a Titan? Anything would be helpful there. Also, if you could walk us through the economic fall through for NVIDIA in terms of what kind of revenues you would expect to see or ASPs NVIDIA would recognize for, let's say, a $300 card and what kind of gross profit dollar fall through we could see from that as well? Thanks a lot.

Yeah, appreciate that. I'd say the GTX 970 is the minimum requirement. It is the minimum requirement. I think Oculus has done a good job setting the minimum requirement. The minimum requirement simply means that anything below it is not going to be a great experience. We're seeing more and more content coming and usually the content gets richer. Content doesn't usually move backwards. If you buy into a GTX 970, it will give you a limited amount of growth over time, but if you buy into a GTX 980 or GTX Titan, you just have a lot more opportunities to enjoy that platform longer as content gets richer. In terms of the economics, our gross margins are higher on GTX 970 and above, higher than our corporate average. We appreciate the adoption and growth in that segment of the marketplace. Ultimately, our focus is to help enable the entire ecosystem of hundreds of companies working on VR to deliver this experience to the marketplace that fulfills the promise of VR. We are concerned that if we don’t do a great job, we ruin a great thing. This is really a great thing. The folks at Oculus, Valve, ourselves and the 230 companies working on VR, we all want to deliver just a fantastic experience. That’s our focus at the moment. Our recommendation, if anyone were to walk up to me and ask me for a recommendation, I would tell them that a GTX 980 Ti would be the baseline and, in my own case, I'll probably have a pair of Titans.

Well, that would be really nice for all of us. On that same theme, if I think about the competitive dynamics into next year, you I think pretty obviously have taken a significant leadership advantage, especially at that enthusiast portion versus AMD. As we go into VR, when you talk to the game developer community, what kinds of things are you doing in order to maintain that advantage? What kinds of things could we see in Pascal that could allow you to extend that advantage?

Sure. First of all, we respect our competitors. We take our competition seriously. You know that as a company we compete pretty intensely. This is a company that has seen a lot of competition over the years. We take competition seriously. However, I think it's also clear that our business and our business model and our strategy is completely different than AMD and the PC graphics chip company we used to be a long time ago. Our company is on a different trajectory. Our approach to building products is different. Our approach to go-to-market is different. Our approach in engaging the ecosystem is different. I can't imagine a more different company. We always alert and paranoid about our position and we don't take our position for granted. We have our own work to do and we have our own platform to do and we have our own ecosystem to care for and nurture and push forward. I appreciate the question, but I can't imagine a more different company.

Great, thanks a lot.

Thank you. Congrats on the solid quarter. This is Jee for Ross Seymore. Just going back to the guidance, given that there's a slate of AAA-rated games coming out this quarter, I guess what's the reason behind the conservatism behind the expectations for gaming, or is that growth expected to be offset by other segments?

Well, our guidance is the best guidance we think we should provide at this moment and we'll see how the market plays out. We're not in control of the market, as you know. At the foundational level, we have multiple growth drivers. Some of them exceed our expectations and some of them don't, but that's one of the benefits of our business model today. We have multiple growth drivers in the company. PC gaming is a growth driver. Artificial intelligence is a growth driver. VR is a growth driver. Autonomous cars is a growth driver. I really don’t know of that many companies in the world working on these important dynamics in the industry that is going to shape the future of computing. We have multiple growth drivers in our company that are fundamentally sound and powerful. However, we'll see how they turn out at the end of the quarter and I remain optimistic. Our guidance is enthusiastic, and we'll just have to see how it turns out.

Thank you. And for my follow-up. It looks like some of the revenue segmentations for the end markets was changed a little bit this quarter with the new classification of Quadro and data center or pro-visualization and data center versus enterprise and HPC cloud. What was behind the change? Is it related to the unifying of the data center strategy?

Yes, that's right. That's exactly right. The reason we decided to classify it a little differently is that I changed the architecture of it. We used to have several different product lines that were planned differently. We've decided to unify all of our data center product lines and abstract the software from it. From this one unified platform, Tesla, Tesla is our accelerated computing platform. On top of it are software stacks that allow us to solve a problem for a particular marketplace. We have our entire software stack for HPC, a rich software stack for enterprise virtualization, we call GRID. We have a software stack that is incredible for providing cloud gaming around the world; that's called GeForce NOW. Then we are going to have software stacks for hyperscale. Each one of these software stacks are built on top of one common architecture, one common platform called Tesla Accelerated Computing. As a result, we decided to unify all of that and report it simply as data centers, which will be easier to understand.

Okay. Thank you.

Thank you for taking my question. Solid job on the quarterly execution. On the strong growth in the gaming segment; blockbuster games, and eSports, big drivers, but there appears to be another potential driver, and that is the shift from console to PC gaming here in the U.S. I think Jen-Hsun, you've always characterized the U.S. as a console-centric market, but this seems to be changing pretty rapidly with high schoolers and some of the older demographics here in the U.S. I don't know what the stats are, but thus far I think it's been a pretty big untapped market. So how much of this penetration is contributing to the growth in your gaming business?

Well, Harlan, first of all, thank you. There is no question that the U.S. is seeing growth in PC gaming. The fundamental reason is really quite singular, eSports. eSports is the new social platform. It's the new digital playground. You go hang out with your friends. It's the new virtual arcade. This is how you hang out with your friends. Not only do you play if you're not playing, you are watching, but at all times you are talking to your friends. This is a big deal. This isn’t something you can do easily on mobile or on TV. This platform benefits from having a PC. eSports is probably, if I had to tag it on one thing, eSports would probably be it. The growth of eSports, as you know, not only is the gamers enjoying eSports and growing faster than anybody would have expected, and I think the reason for that is because it's a social platform, and therefore, it grows not linearly but arguably exponentially. There are also viewerships, and now all kinds of businesses are cropping up to help people enjoy watching eSports and sharing the moment with other people. This is real and quite exciting.

Great. Thank you. And congratulations, again. In terms of the buyback for next year, can you talk about your domestic cash position? Would you need to do anything synthetic to achieve that buyback?

Yeah, thanks, Joe, for the question. We have an outlook for our capital return program next year to return $1 billion, as well as the increase in the dividend that will be in effect there. We continue to look at our overall total cash balance and also our U.S. cash balance. We still receive cash flow from both the combination of our international operations and our U.S. operations, which will take us through the good part of fiscal year 2017 in probably the whole amount. If some need comes, we have several options to think about how we would fund that and we’ll look at that at that time, if necessary.

Great, thank you. I was under the impression from our conversations during the quarter that there was a 14-week quarter in January. Is that the case? And if so, does that have any impact on the financials?

Yes, this fiscal year, we do have an extra week. We have incorporated our best estimates at this time for the fourth quarter incorporating that extra week.

A powerful quarter, congratulations, everyone. Jen-Hsun, I know you touched upon the challenges in the data center for Moore's Law and that driving opportunity for you. But what are your own big picture thoughts on the implications of Moore's Law for your core GPU business? Then associated with that, if I think about the increasing challenges of Moore's Law as well as the power of some of the big OEMs and it can be either on the hardware or on the cloud side, and many of them are both, how does that not mean you need to have some sort of licensing business or strategic partnership business in order to move your technology forward? Thanks.

Yeah. Alex, first of all, thanks for the question. Semiconductors and chips are really important to us. It's a lot easier if the world just gets better without anybody having to work for it, i.e., Moore's Law. If transistors get faster every year, they get more morbid and they get cheaper, all at the same time, it would surely make for an easier ride. However, accelerated computing is a very different beast. The way accelerated computing works is that some people have characterized NVIDIA's growth and performance to Moore's Law squared, and I don't think it's much off from that actually. We change our GPU underneath, the software stack on top, the algorithms we innovate on top of that. Our focus is not just on support; it’s on targeting specific applications. Our business model and strategy have led us, today, to target fields like visual computing where we’ve become quite large. Our primary focus today is through product sales. We changed our business model to move forward and engage the vertical markets directly.

Great, thanks.

Hi, Jen-Hsun. One question on the gaming notebooks. You indicated that now you have the same desktop graphics and notebooks. My question is with the VR coming in and notebooks, graphics getting more powerful, do you see that it could shift from desktops to notebooks, your product mix? And if so, do you think that that mix is less favorable or similarly favorable to you versus desktop graphics?

I guess I would say I don't care. I would say that I don't care and the reason for that is these desktop GPUs are high end. They are all GeForce platforms to us. The most important thing is that we would like the games to have higher graphics production value, the AAA games. I would like eSports to continue to grow. I would like developing countries to have broadband. I would like VR to be incredible. Those are the important drivers and however a gamer would like to enjoy GeForce, we would love to welcome him.

Yes. Thank you very much. Jen-Hsun, you had mentioned earlier in the call in response to a question that you are not fully penetrated in gaming with the eSports phenomenon and more broadband coming to emerging markets. Could you help us understand and break down the growth between unit growth or overall installed base growth versus upgrade sales from the current install base? I guess what I'm trying to get at is the growth between that and what's the upgrade cycle you see when someone says they bought a high-end GPU from you a year ago? What’s the upgrade cycle in links and what are the real drivers that trigger that upgrade cycle for someone that is already engaged with you in your install base? Thanks.

So, one of the most important drivers is the combination of really, really high production value games and our install base, because our install base are already gamers, our installed base are already GeForce customers. When awesome new games come out, our install base wants to upgrade their GPUs to support it. If you look at our install base and take a look at our mainstream GTX, let’s pick the GTX 950. The GTX 950 class install base represents probably less than a quarter of our install base; another way of thinking about it is that three quarters of our install base really needs to get upgraded so they can enjoy today’s exciting AAA titles. That alone has really great growth opportunity. Not to mention the other factors we started with which was eSports and developing countries having broadband, and what we hoped to be a big driver in the coming year with VR. The install base is a big growth driver for us. Okay?

Yeah, thank you, and congratulations on good results. A lot of this has been talked about before, but just wanted to get a sense of kind of the normalized revenue growth rate, at least as much as you can discuss with some of the headwinds that you experienced this year, potentially abating next year, mainly the PC OEM business, it might not decline this fast or at the same rate as it did this year. I was wondering if you could maybe discuss a little bit about more of a normalized growth rate. It seems like this year you are going to end up about 5% growth given the fact that PCs have fallen pretty significantly. So any color there would be helpful?

I appreciate it. The recognition that our PC OEM or our OEM business has largely, the decline has abated. We're selective about the OEM projects, thoughtful about our strategic partners. My expectation is that the PC OEM business and the Tegra OEM business is going to be here. But it's just not a huge focus; it's not where we expect our growth drivers to come from. Our major growth drivers are the four that we've talked about several times during the call. In terms of growth rate, it may take a couple years for us to have some kind of feeling for growth rate. The reason is this: Any time in the history of our company we have had a brand new business model. This new business model leverages one architecture, but instead of serving OEMs, we largely engage the vertical markets ourselves. It’s a vertical strategy. Our engagement in this market is really accelerating. We have multiple growth drivers that are large markets. The automotive market is going to be large. Accelerated data center is going to be large. AI is going to be large, VR is going to be large. We are engaging large markets and have multiple growth drivers. I hope to see growth continue to surprise.

Thanks.

Thank you very much. In your gaming GPUs, do you have any estimate of what market share you currently control and how that's been changing over the last few quarters?

Well, David, thanks for the question. We don't really control anything. Our products are purchased by our gamers freely. It's not a matter of control. We have to earn the love and loyalty of our customers every single day. While we used to be a PC OEM company, an OEM company where that responsibility lies on our OEMs, today that responsibility lies with us. We take it seriously. We have 70 million gamers that are connected with us. I hope we add value to them on a continuous basis and that we earn their trust over time. The beauty of this model, of course, it's much, much harder. Much greater understanding of the software stack is needed. It’s much more complicated than before. Our main goal is to deliver the promises of our products to customers and we hope that through that, we develop their loyalty over time.

Yes, thanks. A question for Colette here. Lots of interesting opportunities and investments here. Could you update us in terms of where we are in terms of identifying and implementing the operating synergies? I think at the Analyst Day you talked about some software synergies unifying the silicon architecture. Thanks.

Thanks for the question. Our focus over the last several years has really been around one single unified architecture and investment. That's really helped us to make the appropriate investments we need in these four different markets we are going after, allowing us also to really focus on a single architecture that allows the efficiencies we’re seeing in our operating expenses. That's still going to be our focus as we move into the current quarter Q4 and as we move forward into fiscal year 2017, as our financial performance is very important to us and improving that financial performance in the short term.

Hey, guys. Let me echo my congrats on a really strong quarter here. The current leader in ADAS, autonomous driving I guess is where they are going with this; they're running their platform off of what's essentially like a $5 microcontroller and they really talk down the importance of hardware. So, Jen-Hsun, perhaps if I could get your idea on what you think this hardware platform will look like for these future generations that you’re talking about.

Yeah, Chris, I appreciate that. If we can get a $5 microcontroller to drive, it’s just a matter of software, driving is one of the most complicated things we do. I think humans could be replaced in just about everything we do with a $3 microcontroller. It seems unlikely. I agree that smart microcontrollers could be used to enhance smart cameras for computer vision; I think that for ADAS it is really, really good. But autonomous driving is a whole different thing. Autonomous driving hasn't really advanced into the market yet; there are early versions but architectures that enable fully autonomous driving have yet to arrive. We really have a lot of capability in this area. It’s going to require sensor fusion, a lot of software. It’s going to require AI and the ability to support an open computing platform that NVIDIA is built upon. If you look at our accelerated computing platform, it’s programmed by many people in the world and this is a real advantage because car companies need the ability to develop their own AI networks for their driving experiences. The new era of autonomous cars and vehicles is a great opportunity for us. There’s still lots to do.

Hey. Thanks for taking my question. Congrats on the quarter. I just wanted to circle back on the automotive part of the business. I know you've done a lot of work with the VW Group, specifically with Audi and Porsche. I was wondering if you have seen any of the impact from the potential R&D clawbacks or any sort of comments on what you are seeing out of Germany would be helpful? Thanks.

The area where we work with VW and Porsche and Audi are really about advancing the state-of-the-art in their cars. Our work is really about innovation and the technology they want to be the world's first in bringing to market. This is a company that believes technology leadership and moving towards the software-defined car and driver autonomous driving and electric vehicles is a good thing. Our work with them is along those lines; if anything at all, we see a significant heightened desire to move the company forward and achieve great things.

Thanks everybody for joining us today. There are several things that we wanted to say. Our new business model transition approach to market is working and accelerating. We talked about our multiple growth market drivers, PC gaming, data center, and automotive are all growing nicely. We're working at the center of some big developments shaping the future of computing—AI, VR, accelerated cloud computing, and autonomous driving cars; these are really exciting things. We're doing that with one singular leveraged investment around one architecture and that allows us to bring the might of this company and the expertise that we have to help solve some of the problems that I have mentioned. NVIDIA is the world leader in visual computing and it's becoming more important than ever in a growing number of industries. Our strategy is to leverage this one core investment to four growth markets, gaming, professional visualization, data center, and auto, and is delivering good results and gaining momentum. Our goal is to balance those investments to capture the enormous opportunity ahead while maintaining a keen focus on improving near-term financial performance. Thank you all for joining us today.