輝達 (NVDA) 2011 Q3 法說會逐字稿

Good afternoon and thank you for holding.

I would now like to turn the call over to Michael Hara, Vice President, Investor Relations.

Thank you.

Sir, you may begin.

Thanks, Will.

Good afternoon, and welcome to NVIDIA's conference call for the third quarter of fiscal 2011.

With me on the call today from NVIDIA are Jen-Hsun Huang, President and Chief Executive Officer, and David White, Chief Financial Officer.

After our prepared remarks, we will open the call up to a question-and-answer session.

Please limit yourself to one initial question with one follow-up question.

Before we begin, I would 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 November 18, 2010, and the webcast will be available for replay until our conference call, when we discuss our financial results for the fourth quarter of fiscal 2011.

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 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-Q for the fiscal period ended August 1, 2010, 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, November 11, 2010, 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 John Petty Research.

With that, let's begin.

We have restored our speed of execution.

Since March, we've built seven new GPUs, all of which with the production with First Silicon.

The Fermi architecture is now in every segment of our desktop, notebook, and workstation product lines.

We are seeing solid gains in our core GeForce and Quadro businesses, and we are recapturing our share.

In the GeForce business, we have extended our DX11 technology leadership and the results show.

Our desktop stand-alone GPU share grew from 55% in Q2 to 59% in Q3.

For budget-conscious gamers, we introduced the GeForce GTS 450, which is designed to deliver the most advanced GPU and gaming experience to the important $99 add-in card segment.

According to the latest data from Steam, a broad serving of hardware used by PC gamers today, 56% of gamers play at resolutions between 1600 by 1050 and 1280 by 1024, resolutions at which the GTS 450 excels.

The GTS 450 follows one of our most successful GPUs, the GTS 250.

In October, we launched the GeForce GT 430.

It is designed to deliver unparalleled experience in photo and video editing, Blu-ray 3D, and the next generation of GPU accelerated web browsers and web content.

The GT 430 is targeted at the $79 add-in card segment and is the perfect first upgrade from integrated graphics.

And this week, we launched the GeForce GTX 580, the world's fastest DX11 GPU.

The GTX 580 is based on the second generation of Fermi architecture, engineered from the transitional level up in just seven short months, and it raises the bar of the fastest and now quietest GPU of its class.

The combination of the innovative vapor chamber and thermal solution and new architectural enhancements increase power efficiency and performance over previous generations.

The GTX 580 delivers up to 35% more performance [per watt], and raw performance is up to 30% greater than that GTX 480 it replaces.

For DX11 games, the GTX 580 is up to 160% faster than the closest competitive product.

For previous DX9 and 10 games, the GTX 580 is faster by up to 62%.

In notebooks, we launched during the quarter a complete family of GeForce 400 MGPUs, with Optimus technology.

Lenovo and Dell have already shipped and more OEMs will follow.

With our new Fermi notebook GPUs, we have significantly increased our design win share.

Starting in Q4 with the Sandy Bridge ramp, we expect to take our notebook share to our all-time highs.

In our Quadro Professional business, the update of the Fermi architecture continues to be strong.

In October, we launched the Quadro 2000 midrange GPU and the Quadro 600 entry-level GPU.

These new professional graphic solutions with the computational and visualization benefits are the breakthrough Fermi architecture within the reach of all engineers, designers, and animators.

The Quadro 2000 delivers 1.5 times the geometry performance of its predecessor, for dramatically higher performance across leading CAD and DDCC applications such as SolidWorks and Autodesk 3dS Max.

The Quadro 600 is an economical half-height solution that features the industry's best performance per watt, for applications such as Autodesk AutoCAD 2011.

As OEMs and software ISVs complete their qualifications on these latest Quadro's, we believe we are attracting historical peak revenue, even in less robust economic conditions.

The adoption of our Tesla supercomputer GPU continues to grow.

Last month, the new Tianhe-1A supercomputer was revealed at HPC 2010 China.

It's capable of over 2.5 petaflops, making it the fastest computer in the world.

The system is located at the National Supercomputing Center in the city of Tianjin.

Tianhe-1A is powered by 7000 Tesla M2050 GPUs.

Had this system been built with CPU's alone, it would have required more than 50,000 CPUs, twice as much floorspace, and three times the electrical power to deliver the same performance.

The difference in power consumption is enough to provide electricity for over 5,000 homes.

We're thrilled to see supercomputing entering its third era.

Created in the 1960s to enable military research, supercomputers were embraced by scientists to simulate models of the world.

The creation of our Tesla parallel processing GPU has come at a time when companies in a wide range of industries are in desperate need of faster supercomputers to accelerate product development.

Tom Dunning, who directs the NTSA and the Institute for Advanced Computing Applications and Technology at the University of Illinois in Urbana-Champaign, put it succinctly when he stated, quote, "What we are really seeing in the efforts in China, as well as the ones we have in the US, is that GPUs are what the future will look like.

What we are seeing is the beginning of something that's going to be happening all over the world." -- end quote.

Supercomputing is no longer a niche; it's a huge growth business.

Another highlight of the quarter was our second GPU Technology Conference held in mid-September.

It was a huge success that underscores the momentum building around parallel computing.

Leaders from national laboratories worldwide, as well as Fortune 500 companies, startups, and the universities, presented more than 280 sessions to more than 2,000 attendees from over 40 countries.

Compared with last year, we had five times the number of top submissions and twice the number of sessions.

The [Excess Scale] Report stated, quote, "Absolutely one of the best and most important conferences in the technology and advanced computing sector." -- end quote.

Heading into Q3, our focus was to restore our speed and excellence in execution, and to drive our media growth opportunities to ramp Fermi-based GeForce and Quadro GPUs and regain GPU leadership and share.

With our basic GPU business now back in growth mode, we can further accelerate our growth with Tesla and Tegra in the coming quarters, and position NVIDIA at the center of the exciting revolution happening with cloud and mobile computing.

With that, let me hand the call over to David.

Thanks, Mike.

Revenue for the third quarter was $844 million, up 4% sequentially.

Gross margin was 46.5%, at the low end of our range, primarily as a result of product mix being slightly less favorable than we had expected at the beginning of the quarter.

OpEx was down to $288 million, principally as a result of our significant improvement in execution of our new product introductions.

And most importantly, we've taken the last seven GPUs from [ASUR and One] Silicon directly into production.

Avoiding the need for re-spend significantly helps to keep our OpEx flat, while dramatically improving our time to market.

Combined, these generated a GAAP net income of $84.9 million or $0.15 per diluted share.

Our focus on inventory reduction continued and tight management enabled us to work it down 13% during the quarter to $378 million.

We achieved this while simultaneously reducing channel inventory slightly.

Revenue per GPU grew slightly and was up nearly 6%.

Total revenue for our professional business was down slightly as a result of the computing segment being down.

Revenue for the computing segment varies from quarter to quarter because it can currently be influenced by a small number of large orders.

As we successfully grow into more vertical segments, this revenue stream should smooth out.

Our customer business was -- consumer business was up slightly, driven by a strong showing in our embedded business.

Tegra revenue should begin to ramp in Q4, as our customers begin building their pre-launch inventory.

Our outlook for the fourth quarter of fiscal 2011 is as follows -- revenue is expected to be up 3% to 5% from the third quarter; GAAP gross margin is expected to be flat; GAAP operating expenses are expected to be approximately $300 million; GAAP tax rate is expected to be 18% to 20%.

That concludes our prepared remarks.

We'll now take questions.

(Operator Instructions).

Mike McConnell, Pacific Crest Securities.

Looking at Q4, could you talk about the drivers and the segment drivers behind the guidance for revenue growth of 3% to 5%, please?

Well, first of all, I think we should all be -- continue to be, I guess, conservative in how we look at the marketplace.

As you know, it's still a little choppy out there.

But there's several things that are working in our favor.

The first is our Fermi GPU is going into the Sandy Bridge announcement.

A lot of people have misunderstood the way the GPU market works and the contribution that a GPU has that makes to premium experience notebooks.

And so whereas Sandy Bridge is a far better CPU and a better GPU than previous generation, the discreet GPUs that are connected to it have made enormous strides as well.

And so our expectation with Sandy Bridge is that our market share is going to go up substantially; the adoption is going to be as good as before; and in fact, the adoption of discreet GPUs and notebooks has been growing over the last several quarters, and our expectation is that, going forward, that trend will continue.

And because Sandy Bridge is going to be such a great processor and a great -- such a great notebook, our expectation is that the Sandy Bridge transition is going to be real exciting.

So that's number one.

Number two, we're going to start building into our phone and tablet design wins.

And although they're a little bit behind, the work that Google and Andy Rubin's team is doing at Google is just really amazing.

I mean, this is clearly a world-class engineering team.

And they're building a magical product.

So I'm looking forward to the next generation of Android phones and tablets that are coming out.

And I think it's going to really, really surprise people and delight consumers everywhere.

And so although it's a little bit behind, I think it's going to be completely fabulous.

Now it's going to take something absolutely great to compete against the iPad and the iPhone, as we all know.

Because both of those devices are quite amazing devices.

And so it's going to take something that is truly remarkable to compete against it.

And I think Andy and his team, and all of our engineers here working with them, and all of our partners around the world working on it, are going to absolutely deliver.

And so that's the second thing.

And Tesla is going to continue to gain momentum.

The number of projects we're working on around the world is growing.

And so we're going to continue to see success there.

There's quite a bit of announcements coming next week at the Supercomputing Conference in New Orleans, so I won't leak any of that, but we're going to have a pretty exciting show next week.

And looking at next year on Tegra, what quarter -- or what's the timeframe we should think about where you'll start to see your Tegra revenue cross over and actually become larger than your chipset revenue?

How should we think about that?

Q1.

Q1.

Okay.

Thank you.

Ambrish Srivastava, BMO.

Just a question on the mix, Jen-Hsun.

The thinking was that as Fermi becomes a larger portion of the mix -- I'm just trying to understand the margin guidance.

Shouldn't that be trending up?

Or is it because the PSP or the professional business is going to be down Q over Q?

And then I had a follow-up as well.

There were some mix -- mixing going on this quarter in terms of high-end versus low-end, professional versus consumer.

But we were coming out of a rather complicated Q2, as you recall.

So I'm delighted with the way things worked out in Q3.

And we're just going to have to improve it from here on out.

So I wouldn't read too much into that.

The important thing to realize is that we're going to recapture a lot of the high-end business that we had seated because we were late to market with Fermi.

And we're going to capture that back and we'll continue to capture that back.

Okay, that's fair (multiple speakers) --

And as we ramp more and more -- more of our mix goes to Quadro and Tesla, both of which are very high margins, that's going to continue to lift our margins.

And of course, the Tegra business has a much longer design win cycle and the amount of contribution of software to the Tegra business is just incredible.

You're building an entire computer on a chip.

And if the computers are as sophisticated as what it's going to take to compete against an iPad or an iPhone, the amount of software that you put into your product line is going to have to increase dramatically.

And so the margins were like that.

So as we continue to increase the mix of these three products, the Tesla product, the Tegra product, one goes into the cloud and one goes into mobile devices and the Quadro product, our margins will continue to increase.

So my sense is that that's the direction that we're headed and there's nothing that tells me otherwise at this point.

Okay.

So my follow-up, just to stick with margins -- there's a lot of concern about pricing.

And now that you have caught up, and potentially incrementally currently better than the existing competitive solution, is there a place where [they'll] be looking at that will drive down GPU -- core GPU margins down, and then Tesla being just the nature of the beast, being very lumpy, how should we just think about that?

Because according to my math, and I could be way off, Tesla is down roughly 30% Q over Q.

Well, the Tesla product is lumpy because, frankly, until now, we've been really the only sales force out there.

And NVIDIA doesn't have a very large enterprise sales force.

But now that starting this last quarter, towards the end of the quarter, Dell, and HP, and IBM all have their systems that they've designed with the Tesla GPU in it, we are now engaging, or if you will, their sales force is engaging into the GPU computing ecosystem.

And so with thousands and thousands of sales and marketing people from the three largest enterprise computing companies in the world selling Tesla now, I think we're in a much better position to gain much smoother traction.

And so the way I think about Tesla up to this point is that we've been doing mostly evangelism, mostly lighthouse design win targeting; and with the momentum that we're seeing, even with that, we're now able to deploy the resources of the OEMs to go after it.

And so I hope that and I believe that with a larger sales force supporting and engaging customers, like all of the big pharmas around the world, all of the supercomputing centers around the world, that all want to advance research; oil and gas companies around the world and finance companies all around the world -- now having a much broader coverage, I think, where you're going to see a smoother ramp and, hopefully, a much more aggressive ramp.

So that's one.

I think the -- with respect to our core or consumer GPU business, you mentioned catch-up.

Frankly, we caught up with Fermi.

We were late with Fermi, but we caught up with Fermi.

We simply pulled further ahead.

You know we (multiple speakers) --

(multiple speakers) So Jen-Hsun, I did say incrementally better.

So, I did give you that credit.

(laughter)

(laughter) No, I wasn't trying to mince words, but you know how prideful we are about the work that we do.

And so I think with GTX 580, whereas we typically launch a new generation each year, because we fell behind with GTX 480, the team has really hustled and brought the next one out very shortly after, based on our second-generation Fermi architecture.

And as a result, we improved our performance per watt by almost 40%.

I mean, that's a huge leap -- using the same exact process.

So as a result, 580 is faster, it's lower power, it's quieter -- I mean, it's just a wonderful product.

And so I think in that segment, winning wins.

And so, this is not a market particularly in the -- from $99 above, this is not -- those aren't markets where lower prices wins the day.

If somebody is willing to spend $199 for a gaming card, they're going to spend $199 to buy the best gaming card.

Where GeForce has a real advantage is across the board on traditional games; but GTX, all of our GTX family has such an amazing DX11 performance compared to the competition.

And Tessellation is the essence of DX11 in the modern generation of games.

And so I think we have, at this point, unambiguously the best architecture.

And we also now unambiguously have the best implementation.

So we just have to market our advantage and I think that gamers will respond.

Okay.

Thanks.

Good luck, guys.

Tim Luke, Barclays Capital.

Jen-Hsun, as you observe the market and you look at the new product traction that you have, and as Tegra gets to begin to ramp, from here -- are you anticipating that you would see sequential improvements over the next few quarters, given that it seems like the computing and graphics market was somewhat soft and maybe now stabilizing to improving?

So, that's my question.

Thank you.

Yes, I appreciate that.

If you look at what we've been trying to do over the last several years, we've been trying to extend our leadership in GPUs and continue to build that base on a GPU business, the core GPU business, which is PCs and workstations.

But in the process, build two new initiatives that we believe was going to be really important growth initiatives.

One of them was GPU for high-performance computing or another way to say it is GPU in the cloud.

Putting GPU in servers available to -- as a computational resource for mobile devices or computers was a really important initiative for us.

And that turned into what we call Tesla.

The second initiative was mobile computing, which is what we call Tegra.

These two products and these two directions of computing we believe was going to be the future of computing.

And along the way, a recession came along and we didn't execute well on Fermi and we tripped ourselves up a little bit.

But we gained our footing again.

Q2 was -- we got it out of our system.

And since then, we've executed on seven Fermi's -- seven Fermi generation GPUs, all of them went to production with First Silicon.

Every single one of them went from tape-out to production release at record times compared to before.

The velocity of the Company has really, really amped up.

And so I think our core business of the GeForce and the Quadro business are really, really in great footing now.

And just as we've gained that footing, Tesla and Tegra are both ramping into production and helping us get into these two really important growth markets.

So I could see the traction now and we know what we're working on with respect to customer projects.

And we have a good feeling for when they go to production.

So at this point, we really feel like we're back to growing.

And more importantly than just back to growing, we're growing into the right markets, which is the future of computing.

So, as a follow-on, obviously you only have a certain amount of visibility, but given that seasonality has been a bit off some of the markets recently and inventory seems to be being cleared, would it be the case that you wouldn't necessarily see the usual sequential decline in April, given that you've got the new products coming on?

And maybe just as a clarification in the consumer area, could you just share with us what Tegra might be as a run rate for the January quarter?

I know you've shared some numbers in the past and any clarification there would be helpful.

The January quarter is -- let me say what I just said earlier, which is -- we know at this point that the projects that we're working on, particularly in the tablets, has taken longer than we expected.

And it's taken longer than we expected by a few months.

But the important thing is that you can't just build a tablet.

You can't just put an operating system on a tablet and hope that -- on a piece of glass and hope that you're going to compete against the iPad.

The iPad is a wonderful product.

And if you're going to give that wonderful product a run for its money, you better build something absolutely exquisite.

And so whether it's the quality of the work of the craftsmanship that our teams are working on, or the capabilities of these devices, they have to be absolutely groundbreaking or why would anybody come to buy them?

So I think that the extra time that was necessary to build these devices, and build the operating system and all the applications and the system software necessary to do it -- and obviously, we're not going to talk about what they are right now, but they're going to be absolutely magical.

And so I think that that's the important thing to realize.

And I think that although we're a little bit late, the market potential is so huge, and this is the future of computing.

This is the second device into the future of computing.

And so I think we're pretty well-positioned.

And that's why I see a pretty good ramp through Q1, through the next year.

Thanks, guys.

Yes, thanks a lot, Tim.

Patrick Wang, Wedbush Morgan.

Jen-Hsun, earlier you had said that you expect the crossover between Tegra and chipset to be potentially the first quarter.

I'm just curious if you could maybe talk about what's going on -- what you're seeing in that chipset market.

And I guess there was an expectation that that was going to come down in the first half of next year, but it seems like Apple has revived that a bit.

If you could just talk about that a bit.

Thanks.

Yes.

It will come down but it won't come down nearly at the levels that most people thought.

And it is the case that the new Mac Air's are pretty amazing.

And we're really proud to have been selected to be inside them.

And if you haven't had a chance to try them, they're really magical computers.

And so I (multiple speakers) --

I've got one.

Yes, so, that's great, thank you.

I appreciate it.

(laughter) Thank you.

I try to help out, you know.

(laughter) If you could spread the word, tell your friends -- we're going to postpone that delay -- that transition.

So I think -- we built with Apple a really important new GPU with integrated chipset and it's a great chip.

And I'm delighted to see them use it and use it in so many different systems.

But our tablet business and our phone business is going to ramp and it's going to ramp hard.

And there are customers lined up all over the world with a lot of different phones and systems.

And the amount of activity in our Tegra business is just simply too great right now.

So we've got to -- we've just got to hunker down and get these products to market; make sure that we build something really amazing.

And all the engineers that are working on it are working 24/7.

So we just have to build some of the best things we've ever built.

So I think Q1 is going to -- Q1 is when all of -- come beginning of the year is when all of you will see what we've been so busy working on.

And I think you'll be more than surprised and delighted.

Got you.

And I'll throw my vote of confidence behind that.

I'm loving this MacBook; I really think it's terrific.

For my follow-up, I was hoping you can talk about -- you guys had talked about gross margins a bit here -- maybe some of the moving pieces here between Q3 and Q4.

And then, just based on the kind of mix you're talking about for next year with a lot of Tegra, I know it's a pretty good margin product there, if we should think about margins being in excess of 45%, 46% just across the year?

I think it's a foregone conclusion at this point.

Certainly, inside the Company that as we continue to increase the investment into Tesla, Quadro, and Tegra, all of which are nowhere near our current corporate average, that as we continue to drive the growth of those three businesses, the margins will go well above 46%.

So wherever it is currently today, it's completely illogical for us to be near it next year.

Got it.

Thanks so much.

Shawn Webster, Macquarie.

On the GPU business in terms of the detail underlying, can you give us a sense of what your GPU units and pricing did sequentially for Q3?

Shawn, from a unit standpoint, units were up sequentially quarter-over-quarter, particularly as we launched new products, Fermi products into the mainstream segment.

They were up pretty strongly.

Desktop overall was up about 6 points.

From an ASP standpoint, that mix change also brought our ASPs down a little bit, as a result of having a richer mix in the mainstream segment of the market.

So as a result, margins were a little bit less than what we anticipated beginning of the quarter, but the good side of it was that desktop was stronger than what we anticipated.

Right.

And then how big is the chipset business?

Did that grow sequentially in revenues for you?

It did grow sequentially.

Okay.

And then when you look at -- you mentioned that your notebook share gains going into the Sandy Bridge generation could reach or exceed your prior peaks.

Your market share this year has been somewhere in the 45% to 50% range on notebook GPUs.

And if I remember correctly, you had been in the low 70's percent share.

Is that the magnitude of share gains you expect next year in notebook?

I would say that our prior peak was probably in the low 60s.

And this year, our market share was actually lower than what you gave me credit for just now.

So I appreciate the graciousness, but (laughter) --

Thank you for the correction.

-- you know, we -- I didn't do that well this year.

Okay.

So as you look across the landscape, there's a lot of design win activity in the tablet space.

Do you have a sense on what your percent share in the tablet area is as we go into next year, in terms of the design win share?

I would say, just a lot.

And I -- there's a lot of tablets that are going to get built that aren't really going to ship.

And everybody is building tablets because it's just so important; this is such an important part of the future computing market.

This isn't a fad.

It is clear now that touch computing and tablets is a wonderful way for a lot of consumers to enjoy content.

And by being able to connect a wireless keyboard and a mouse, the difference between a tablet and a notebook is pretty marginal.

And so I think that going forward, you're going to see this tablet device being quite disruptive to both notebook as well as entry desktop.

So I think this is an important trend.

And that's why a lot of companies -- every single company I know is working on a tablet.

There's car companies working on tablets.

Consumer electronics companies working on tablets.

Computer companies working on tablets and communications companies working on tablets.

I don't remember the last time in the history of computing where a singular device was being worked on by all of the industries.

The medical industry is working on tablets.

This is a revolutionary form factor and I think it's a foregone conclusion this is going to be probably the largest computing segment.

Now, how much of that share depends on ultimately how delightful all of these tablets that are being built are.

And I'm just so glad, I'm so glad that we decided not to go to market with tablets earlier, but to wait and work and put all of our energy behind the next-generation operating system.

And the team over at Google working on it, Andy Rubin and his team are working 24/7 and they're doing amazing work.

And I'm just really, really delighted that we decided to focus and not spare it on current generation operating systems.

And so I think when you see it in near-future, it will more than delight you.

I think you'll be shocked how wonderful it is.

Great.

Thank you very much.

James Schneider, Goldman Sachs.

Thanks for taking my question.

Going back to the Tegra questions before, I understand the reasons for some of the delays in product push-out and the fact that you expect a pretty strong ramp.

And you've also talked about some expectations in terms of Tegra revenue over the course of the year.

Can you maybe give us an update on where you think this might be on a run rate basis at this time next year, given the delays in the products?

You know, I genuinely don't know.

Is this Jim?

Yes.

Jim, I genuinely don't know.

There's only a couple of things that I do know.

I believe that mobile computing is the future of computing.

I believe that the tablet form factor and touch is not a fad.

And I believe that this particular device, this particular form factor, is likely to be very, very disruptive to the traditional form factors of computing.

And if you gave somebody a choice between a netbook and a tablet, and a wonderful tablet, why would anybody choose a netbook?

And so I think the logic I think at this point is pretty clear and people see the trends.

And how big is it going to be to us?

Depends on how well we execute.

This is a segment that I hope that we can add a lot of value to.

And computing matters here.

And very, very, very low-power computing is what Tegra was designed to do.

This is something that we've been working on, frankly, for five years.

And of the three of the largest computing technology companies in the world that contributes to the PC industry today, we're the only one that has made the fundamental investments to be here.

And now we're really well-positioned to make a contribution to the mobile space with computing technologies like ours.

So I think this is a big opportunity for us and I think it's a big market.

And the rest of it just is going to require a lot of focus and execution on our part.

Understand, thanks.

And just as a follow-up on that, can you talk about whether you -- it sounds like you're expecting tablets to be a bigger contributor from Tegra than smartphones.

And then one -- is that correct?

And then one for David -- OpEx has kind of moved around a little bit.

It's declined here and you talked about some reasons why.

Can you talk about some of the reasons for the increase going into Q4 and how we should think about the profile of OpEx next year?

Thanks.

Our smartphone business is actually larger than our tablet business.

And these smartphones -- are they a small tablet or are they a large phone?

It sounds -- from my perspective, it's kind of hard to say.

It's a five-inch tablet, a seven-inch tablet, a 10-inch tablet -- they're all, to me, tablets.

And so a phone to me conveys a dial-in keyboard and it just doesn't convey the right image to me.

So I talk about them interchangeably, but our smartphone business is likely to be larger.

But in combination, our Tegra business should be quite large.

I mean, it will be one of our largest businesses over time.

And Jim, on the OpEx side, we've talked the last few quarters about bringing expenses down and we've guided roughly in the $300 million range the last couple of quarters.

We had some good fortune this last quarter in the fact that while we planned for heavier NRE activity associated with some of the launches of the products we did in the third quarter, we were fortunate that First Silicon was successful.

So we had some benefits there.

Now, going into the fourth quarter where the NRE activity can be lumpy, dependent upon announcements of products and so forth, but we're expecting some of the NRE to go up in the fourth quarter.

We also have payroll taxes coming up in the next quarter, because our fiscal year goes into the next calendar year where FICA taxes start over.

We also have focal increases that we're planning and have executed on.

So, we're striving to keep our expenses roughly in that range for now and some of the lumpiness is just some of the NRE activities and so forth that we've benefited from.

That's helpful.

Thanks very much.

Arnab Chanda, ROTH Capital.

A couple of things -- one is there's a lot of magic going on here but what exactly are we waiting for?

Is it an operating system update?

Is it [true] availability to really see a ramp in Tegra?

And then I have one question about your R&D.

Obviously, for good reasons your revenues are not changed since the second half '06, but R&D has almost doubled.

You're seeing now real fruition in those products you developed.

Can we assume from here on you're going to get R&D growth slowly versus sales?

Or when can we see that happening?

Thanks.

To call the next-generation devices and operating system update would really be an understatement, I think.

And it would do injustice to the amount of engineering that has gone into building the next generation of tablets.

I mean, Arnab, as I mentioned earlier, the iPad is not your normal device waiting around to -- for somebody to beat it.

I mean, this is an extraordinary device.

And if you want to build something that is desirable, and even more desirable to some other customers, you've got to build something great.

You're not just going to put a phone operating system on a larger display and ship it; you're just not.

And so the amount of engineering that has gone into building the next-generation device is not an operating system upgrade by any -- you won't recognize it, let me just put it that way.

And what the team is shooting for is nothing short of a great experience.

And so I have every confidence in the team.

Andy and his guys are doing amazing work and we're putting all of our best on it.

And, of course, many partners around the world are putting their best on it.

We're going to make sure we build something absolutely, absolutely magical.

So that's job one.

With respect to R&D, you're right that as -- although the PC industry hasn't grown substantially in the high end over the last several years, we continue to invest so that we could build two major initiatives -- one was Tesla, so that we could put GPUs into servers in the cloud.

And one was Tegra, so we could put our GPU and technology into mobile devices.

I think it is clear, surely, I think, to us and I hope to you, that if we hadn't made the investment to put NVIDIA in the middle of cloud computing and mobile computing, and that today we are basically selling PC chips, it would be one of the greatest irresponsible acts of investment that I can imagine.

Here is an incredibly talented organization that knows how to build these things, and for us to be at the epicenter of the next computer revolution is an opportunity too great not to have gone after.

And so we decided to invest heavily to be here.

And I think that over the next year, and we're starting to see it now, the investment was surely worth it.

And that we've made the right investments to be at the growth -- the next growth part of the world, next growth part of the industry going forward.

And so I'm excited about the investment that we've made, frankly.

And I think we've made really important contributions to the computer industry that you'll see in the very short-near future.

Thanks, Jen-Hsun.

Yes.

Thanks a lot, Arnab.

Harlan Sur, JPMorgan.

Thank you for taking my question.

You talked about channel inventories being lower.

This is the second consecutive quarter, so nice job on that front.

Can you guys help quantify where those levels are relative to historical of about -- I think it's about seven to eight weeks?

So, you know there's -- there is not a great deal of science that we can go through the entire channel of inventory and so forth, but our targets have been to keep channel inventory somewhere in the four to six-week range.

And today it's well within that range and we anticipate keeping it in that area.

Okay, great.

Thanks.

And you guys are on the cusp of ramping Tegra here into smartphones and into tablet customers this quarter.

Jen-Hsun, can you just remind us what's next on the road map for Next-Gen Tegra?

When is the team going to introduce this to the market?

I know competitively, we've seen some three core architectures being rolled out by your competitors.

Just curious, what's next for NVIDIA?

Well, if you look at the competition, they -- about a year -- Tegra 2 was ready for production in Q1 of this year.

And obviously, we had to -- in order to build a great device, you need a lot of software to go with it, including operating systems and applications.

And that's taken us pretty much the entire year to go do.

Meanwhile we've -- as the competition is finally catching up to where we were last year -- beginning of the year, we're about to enter into our next generation.

And so I think today is probably too early to talk about -- I'm anxious to sell you some Tegra 2's, so I'm anxious to get everybody excited about Tegra 2's.

And I think they will be very excited about it.

And then shortly after that, we'll introduce Tegra 3.

But having a rhythmic road map once a year is what the Company is focused on and we'll surely deliver that.

Okay.

Thank you very much.

Glen Yeung, Citi.

Hey, Jen-Hsun, just going back to the discreet business, can you just clarify -- you know, AMD at their Analyst Day sort of implied that attach rates in 2011 would not really change from 2010.

What's your perspective on that sort of on an overall basis for the discreet market?

I don't think attach will change.

There's no indication at all that the attach rate is changing.

You know, integrated graphics has improved over the course of the last 10 years, last 15 years -- it's been 15 years, I think -- I think about 11 years, okay.

1999.

Oh, is that right?

Okay, yes.

So, yes, right.

So it's been 11, almost 12 years that integrated graphics has been around.

And every single year, it's gotten many times better than the previous year.

And you almost don't have to try for it to be better every year.

And so I don't expect -- so this year it's better than last year.

And Sandy Bridge is a great processor; I'm excited about it.

I think it's the best processor that Intel has made in several years and it's by far the best processor on the planet.

And so I expect that the refresh cycle associated with Sandy Bridge to be quite exciting, both on the desktop side and on the notebook side.

I surely am.

I'm going to be one of the first in line to go get myself a Sandy Bridge.

So I'm excited about that processor refresh.

But it doesn't change the dynamic -- nothing that we've seen so far has made integrated graphics anything close to what a discreet graphics could (multiple speakers) --

Do you think it might change the dynamic between high-end and low-end attach, i.e., maybe skew the attach rate more towards higher-end GPUs as part of the mix?

Well, if what you mean by that is that the discreet GPU has to be even better in order to be better than integrated graphics, then the answer is yes.

And we know that -- we know now that we've done a really quite a fine job choosing the performance levels of our next generation GPUs, so that it's still substantially better than the Sandy Bridge.

And so my sense is that no dynamic has changed.

So that's one area I would agree with AMD.

Okay.

The one and only.

That was just a joke.

That was a joke.

One other question for you is -- particularly, you guys are in a pretty unique position, inasmuch as you're very well-positioned in tablets.

What's your thought on the potential of cannibalization of tablets into the PC market?

And then sort of as an adjunct to that, what's your expectation for PC growth in 2011?

Well, I think the segment of the market that we serve is stable and hasn't really grown much over the years with respect to consumer PCs.

I mean, we're going to have to discover another great reason for people to grow the high-end PC marketplace.

But our discreet GPU GeForce market has been relatively stable for some time and -- but the growth has all come from low-end notebooks and netbooks and devices like that.

And I think that that market is absolutely going to get cannibalized by the netbook -- by the tablet.

And the reason for that is because the tablet is just much more delightful to use, on the one hand, and if computing is a secondary device anyways, or it's for casual computing, why not use the tablet?

You know, the applications are all served up wonderfully, and the experience is wonderful and responses are beautiful.

It's an exquisite device.

And if you want to use it as an entry device, you can always attach a keyboard to it, a wireless keyboard to it, or a wireless mouse.

And so those things are really easy to solve.

And before you know it, how people look at a tablet, that maybe it is a tablet or a transformable netbook, it's kind of hard to say.

So I think it's completely disruptive.

So your outlook for PCs in 2011?

Well, I don't know where 2011 is going to be, but I'm pretty sure that 2015, the PC industry is going to look very different.

Okay, great.

Thanks.

Craig Berger, FBR Capital Management.

Thanks for taking my question.

I guess, just on the chipset, can you give us an idea of how big that business is still?

Is it kind of in the 15% of revenues range?

And how does that tail off over the year?

And just as a clarification, are you guys still working on any new product in that sector?

Are you sticking with your plan to completely retrench?

Well, let me take the second question first, because I think I know the answer to that one really well.

The answer is no.

We're not building any more chipsets.

We're building SoCs now.

And we're building Tegra SoCs.

And so we're going to take integration to a new level, and that's our focus.

And the second thing is, with respect to the chipset business, it's not grown largely this year because we haven't really been expanding the sales of it.

And with respect to the tail-off, most people expected a tail-off, I guess, come Q1 next year.

But I think the tail-off is going to take a little longer than that for a couple of reasons.

On the AMD side, our AMD chipset remains quite well-positioned.

My sense is that our chipset there will continue to ship throughout next year.

And then the second thing is the MCP89, the latest and the last generation of Intel chipsets that we built, was just a really wonderful piece of engineering.

And the work that we did with Apple was really great and they're going to continue to use that for some time.

So I think that the tail-off is just going to take a little longer than people expected, but I don't know exactly how long.

Is it still about 15% of revenues?

No.

We don't break it out.

Thanks for the detail.

Dan Morris, Oppenheimer.

Calling in for Schafer today.

Just going back to your notebook share comments, could you talk a little bit more about what's driving that?

I mean, is it simply the refresh of the Fermi-based product stack?

Or is it also the separation from the [bum metal] issues?

Well, the answer is actually quite simple.

Our Sandy Bridge GPU is just the best in the world.

And it has Fermi in there.

And we -- this is our second generation Fermi, and so you saw one evidence of it, with GTX 580.

The engineering team just did a wonderful job and their execution is just fabulous.

So that's one.

And the second thing is Optimus technology, it's just too important.

You probably heard about Optimus, but it's this automatic hybrid GPU computing solution.

And it's automatic, it's seamless, it's just delightful.

And no switching buttons, no rebooting; the computer never stalls.

It just -- our GPU turns on automatically whenever you need it.

And whenever you don't need it, it goes to sleep completely; consumes no power.

And so I just think that Optimus is one of the best things we've ever done and it's such great technology, it disappeared.

And I think everybody now -- every notebook company in the world wants to build computers and put a premium experience -- you know, there's the baseline computers where integrated graphics is fine, but you also want to offer premium experience notebooks.

And this is where they put our GeForce in it.

But they still would like to have excellent battery life.

And so with a discreet GPU, our next-generation Fermi-based GPU and also with Optimus, they get the benefit of great experience and very, very long battery life.

So it's sure a non-compromise solution.

And so those are really the two reasons that we've captured so much more share.

Okay, great.

And historically, you've been obviously well-positioned in the high-end and the performance segments, you're really dominating in that space.

You highlighted, in Mike's prepared remarks, the 450 and the 430.

Are you -- is it fair to say that you guys are paying more attention and investing more now in the value in mainstream market, and maybe to provide what I guess -- I'd call maybe a macro bumper for these times when OEMs and consumers de-spec?

We've always offered entry-level GPUs.

The very first entry-level GPU that we offered was called Vanta.

And it had a whopping -- whopping 7 million transistors.

And it took all of our might to build.

So anyways, we built a whole -- we built entry-level GPUs for every architecture, every season.

And we haven't lost focus of our high-end.

You know, Fermi, you could argue that we were too ambitious to focus on the high-end.

We wanted to bring a level of geometry performance to the world that it had never seen.

And we took enormous architectural risk to change our architecture completely, so that the geometry processing capability of our GPU reached a level that really the world had never seen, and the world has not since seen.

And I don't expect that you'll see it from our competition all next year.

And so Fermi is clearly a revolutionary architecture.

And it kind of broke our pick, if you will.

But we gathered ourselves again and the engineering team since then has gone into lightspeed.

And we executed on seven new GPUs, all working the first time, and the velocity of the Company is as fast as ever.

And so I'm really proud of the team for having the willingness to take a chance and innovate like that.

And after the stumble, they picked themselves up and back to rare form again.

Okay, great.

Just maybe to follow-up on that, with this more of a -- I don't know if I'd call it more balanced architectural approach, is there any implications for margins?

Well, Fermi is -- the geometry processing part of Fermi is hard to do, but oftentimes what's hard to do doesn't always have to be big to do.

Things that are breakthroughs don't have to be physically large; they're just complicated.

And so Fermi generates great margins for us.

And if not for Fermi, we wouldn't have the benefit of Tesla and we wouldn't have the benefit of the leadership that we now enjoy in Quadro.

And so you have to attribute the margins that Tesla and Quadro generates to the architectural advantages of our GeForce GPU.

So you can't really separate the two of them.

Thank you.

Ross Seymore, Deutsche Bank.

Just to try to get a little color on that crossover between the chipset side and the Tegra side, I guess starting on the good news side with the Tegra side of things, can you remind us just in a rough range what sort of ASP we'd be looking at, so we can all do our math on the unit side and figure out generally a revenue trajectory?

We've not talked about the ASPs of our Tegra processor, have we?

We have?

What did we tell them?

Generally in the range of $15 to $30.

And the difference -- yes, so the difference is, is on a lower end, it's for smartphones where the volumes are much higher for design win, and for the higher range is for tablets, where each project tends not to generate the level of volumes that a smartphone generates.

So, based on those ASPs, the units that you would have to have to get above the dollars you're talking about -- that math still all works out to get above the chipset side you're talking about?

Oh, yes.

(multiple speakers)

Wow.

(laughter)

Yes, the thing you also have to realize is that the margins -- you know, there's a margin difference between the two as well, right?

So neutrality from a net income standpoint comes at a lower level of revenue dollars with Tegra than it did with MPC.

Got you.

And then just a question on the non-Tesla part of your professional segment.

You said Quadro was up in the quarter.

How are you seeing that business in a general sense?

There's been some choppiness from some of the big enterprise-level OEMs that I'm sure you're aware of.

How is that business going for you?

Any acceleration, slow down, or is it company-specific market share gains kind of mask that off?

Tegra is about to reach their all-time high -- about to.

They're not quite there yet.

And so Quadro -- I'm sorry, Quadro is about to reach their all-time high.

So my expectation is that the Quadro business is going to continue to grow.

And the reason isn't so much just because of we're serving the workstation market; I mean, we're creating new technologies that grows our market.

And so, for example, at GTT, we demonstrated the world's first interactive ray tracer.

And this allows people to do industrial design and styling and walk-throughs and -- in a way that they couldn't conceive of in the past.

And this a cloud computing initiative and so this puts Quadro in the cloud.

So our expectation is not only will you have -- will we continue to benefit from workstation but that continue to adopt Quadro.

And now the people that are going into a new refresh cycle, their workstations are getting pretty old.

Many of the workstations are now three or four years old and it's time to upgrade them.

And the performance difference between what Quadro brings and what's in the field is not 2x or 4x; I mean, it's 10x the performance.

And so there's an upgrade cycle that's always going on.

And then, secondarily, we're going to create new markets and new opportunities for ourselves by putting Quadro into the cloud.

So that's what the [I-rating] initiative is about and that's what we call the reality server, which is a software stack that allows for distributed multi-computer, multi-GPU rendering, serving up to multiple users.

So that particular technology is in the process of being implemented and rolled out to the world.

Okay.

So (multiple speakers) --

Great.

Thank you.

-- we're just trying to create new markets as well.

Our last question comes from the line of Raji Gill with Needham & Company.

Please proceed with your question.

Yes, thanks and congrats on kind of turning the corner.

A follow-up question on the discreet GPU business.

We talked a little bit before, but even if you were to assume, say, Intel's integrated graphics would cannibalize some of the low-end discreet GPU market, would it really have a financial impact on your GPU business?

Maybe just talk a little bit about that, just assuming that that were to happen?

I guess we could theorize about that, but it's not.

You know, there's all kinds of stuff to theorize about, but it's just not happening.

I mean, we know what design wins we have for Sandy Bridge.

And Sandy Bridge is practically in production and we're in production with all these notebooks.

I guess we can fantasize about it but it's just -- it's not.

It's not happening.

We can fantasize about a whole bunch of other stuff but that one is just not happening.

I don't need to fantasize about that because reality is right in front of me.

And just -- David, on the share buyback, you had put an authorization earlier in the year.

Just wondering, any thoughts on that?

You know, we are always looking at our liquidity needs as a company, reevaluating that; but at present, we don't really have any plans for a repurchase.

Thank you.

Okay.

And there are no further questions at this time.

I'll now turn the call over to Jen-Hsun.

Please go ahead.

Thanks, everyone.

We'll look forward to talking to you about our Q4 results.

And ladies and gentlemen, that does conclude the conference call for today.

We thank you for your participation and ask that you please disconnect your lines.

Have a great day, everyone.