輝達 (NVDA) 2011 Q2 法說會逐字稿

Good afternoon.

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.

Good afternoon and welcome to NVIDIA's conference call for the second 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 up the call 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.

It is also being recorded.

A replay of the conference call will be available via the telephone until August 19, 2010, and the webcast will be available for replay until our conference call to discuss our financial results for our third 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 May 2, 2010, and the reports we may file from time to time on Form 8-K filed with the Securities and Exchange Commission.

All of our statements are made as of today, August 12, 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.

Rapidly changing market conditions made Q2 a challenging quarter.

We experienced softness in consumer PC demand in Europe and China, our largest GeForce regions.

Weak consumer market and the weak euro shifted demand away from the premium experienced PC segment that G4 serves.

At the end of Q1 we began selling our new Fermi-based GeForce GTX 480 and 470 GPUs into strong demand and saw healthy in-channel sellout for GeForce products across our lineup.

We entered Q2 with high expectations.

As the quarter progressed, rising memory costs and the weakness of the euro increased in-the-market prices of graphics [at in-carts].

In addition, the growing economic concerns in Europe and China began to create pressure on discretionary spending.

Because discrete GPU attach rates in Europe and China are among the highest in the world and our share in both these regions is also very high, these factors hit our consumer business particularly hard.

Today we also announced an additional charge to cover costs resulting from a weak die packaging materials set in certain GPU and MCP products shipped before July 2008.

This is the same issue we disclosed previously.

We provided more detail in the CFO commentary which we released prior to the call.

Although the end markets remain uncertain, we have exciting new products that can drive revenue and share growth.

We are continuing to regain our GeForce consumer GPU leadership position across the entire product stack.

The GTX 40 has claimed the top spot in the enthusiast segment.

Our focus is now on the gamer's sweet spot, the $199 performance segment.

In July we launched the GTX 460.

GTX 460 uses our second-generation Fermi architecture and delivers the perfect balance of price, performance and power specifically for the $199 segment.

The industry reviews underscore the 460's unrivaled competitive position.

HardOCP (technical difficulty) it performs admirably, it's dead quiet, it tips power, it runs cool and it over-quads like a monster without voltage modification, end quote.

GameStar Hardware explained, quote, NVIDIA is back.

The GTX 460 is fast, really fast, and at the same time enjoyably quiet and has much lower power consumption, end quote.

HardOCP also wrote in the following report, quote, if our earlier articles have not proved it yet, this one surely puts the final nail in the coffin.

GTX 460 1-gig SLI real-world gaming performance is superior to even AMD's fastest high-end single GPU video card in a Crossfire-X configuration, end quote.

AnandTech put it simply, quote, NVIDIA's GeForce GTX 460 -- the $200 King, end quote.

GTX 460 is a showcase of the revolutionary Fermi architecture designed from the ground up for DX11.

We are now ramping the rest of the Fermi GPUs from top to bottom and including notebooks and workstations.

At SIGGRAPH, we launched the era of computational visualization for designers, engineers, researchers and animators with the introduction of our Fermi-class Quadro GPUs and new NVIDIA 3D Vision Pro solution.

Computational visualization combined with the parallel computing power of the Fermi architecture with visualization to enable a breakthrough capability for designers.

Powered by our new Fermi-class Quadro, artists can create photorealistic imageries for print ads, commercials and movies in minutes rather than hours.

Car designers can simultaneously optimize the design for beauty as well as for aerodynamics.

The new family of Quadro GPUs deliver performance that is up to five times faster for graphics and up to eight times faster performance for computational simulation than our previous Quadro generation.

Premium class Quadro is the biggest discontinuity we've introduced since Quadro FX introduced geometry processing over 10 years ago and workstation partners Dell, HP and Lenovo are all racing to engage new markets as well as upgrading a large installed base of Quadro workstations.

Tesla achieved another record quarter and continues to make further inroads into our targeted verticals.

Last week DARPA announced that NVIDIA has been awarded a $25 million research grant (technical difficulty) to address what the agency calls, quote, a crisis in computing.

Today's conventional computers are starting to hit practical limits since they can only do so much computation, given a certain amount of energy.

For DARPA to recognize that NVIDIA should take the lead to discover the computing architecture of the future is a recognition of the impact of GPU computing and NVIDIA's contribution to computing.

Under a four-year contract the team plans to build a new class of exascale supercomputers that are 1000 times more powerful than today's supercomputers, using similar amounts of energy.

Our team includes Cray, Oak Ridge National Laboratory and several top universities.

The enthusiasm for our upcoming GPU Technology Conference in September is another example of the momentum we are seeing behind GPU computing.

Last year we exceeded attendees' expectations.

This year we have four times the number of papers submitted, three times the sponsorship levels and are expecting twice the number of registered attendees.

Industry speakers will discuss the use of CUDA in fields ranging from cloud computing, computer vision, energy exploration, life sciences, medical imaging to finance.

It is very clear now that parallel computing has reached the tipping point and is accelerating discovery in a broad range of important industries.

Finally, in one of our most important initiatives, we need to execute and help our Tegra customers ship their next-generation smartphones, tablets, netbooks and cards.

These devices will be the industry's first to offer Tegra's dual-core CPUs and HD resolution graphics and video.

We will begin the production shipments for some of our customers' key projects this quarter.

We are in the home stretch.

Regaining GPU leadership and share, wrapping Fermi-based Quadro, continued adoption of Tesla and shipping Tegra customers -- these are the immediate growth opportunities we are focused on.

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

Revenue was $811.2 million, down 19% sequentially.

GAAP gross margin was 16.6%.

GAAP OpEx was $309.5 million, and GAAP net loss was $141 million or $0.25 per share.

Reflected in these results is the net charge of $193.9 million related to the weak die/packaging material set.

Non-GAAP net income, which excludes the net charge, was $20.1 million or $0.03 per diluted share.

GAAP gross margin of 16.6% for the second quarter was unfavorably impacted by two significant events.

First, GAAP gross profit was impacted by the aforementioned charge.

Second, the market shift in our consumer GPU business resulted in excess inventory of certain primarily older generation products.

As a consequence, our second-quarter results include charges for a large inventory write-down.

Excluding each of these items, gross margin would have exceeded our outlook.

The inventories at the end of the quarter were $434.2 million, up 11.9% over the prior quarter.

Notwithstanding the inventory write-down, inventory was up as a result of lower actual revenue than was planned early in the first quarter, when we made wafer start commitments.

Our manufacturing cycle time is approximately four months.

We made appropriate adjustments to our build plans over the course of the second quarter, but we don't expect these corrections to have a meaningful impact on our inventory levels until the fourth quarter.

Net of the inventory write-down we took in the second quarter, we believe the inventory mix in our current pipeline is aligned with current market demand.

Revenue for the GPU business was down 29.5% due to reasons discussed earlier.

We do not expect the market conditions to change dramatically, but we expect our competitive position to improve and regain some segment share.

We launched our new GTX 460 into the $199 performance segment, and it has received outstanding industry reviews and continues to sell well and gain share.

Revenue for our Quadro professional solutions and Tesla computing solutions represented 26.5% of total revenue and was up 13.4% over the prior quarter.

Tesla achieved another record quarter.

Revenue for the consumer products business was up 46.5% over the first quarter.

Substantially all the increase was attributable to Tegra.

Cash and cash equivalents and marketable securities at the end of the quarter were up $1.77 billion, up modestly from the first quarter.

Our outlook for the third quarter of fiscal 2011 -- revenue is expected to be up 3% to 5%.

GAAP gross margin is expected to increase to 46.5% to 47.5%.

GAAP operating expenses are expected to be approximately $300 million.

Our tax rate is up 17% to 19%.

That concludes our prepared remarks, and at this point we will now take questions.

(Operator instructions) Raji Gill, Needham & Co.

Two questions, one on the revenue guidance of up to 3% to 5%.

Maybe you could provide a little bit more color where you see that guidance, given the weakness in the PC supply chain.

Also on the margins, significantly -- the guidance is significantly higher.

What gives you confidence that you will be able to get the margins back up to the level?

Would you not expect price competition to come back into the market?

Maybe just some more details on the gross margin guidance would be helpful.

So we are assuming the consumer PC market to remain uncertain, and all of the growth that we talked about in Mike's previous comments are the ones that we are focused on.

The GeForce GTX 460 goes into a segment of the marketplace that's really about gamers and gamers come out to buy new graphics cards when there are new great games.

This last month we saw the launch of StarCraft II, which is arguably one of the most important video games in the entire PC industry, with some 12 million plus around the world.

And StarCraft II, for all the people that are on the phone that knows about it obviously hasn't been refreshed since 1998.

I think that's over 12 years or something like that.

So the market has been extremely enthusiastic about it, and the sell-through has been really fabulous.

And so I think this is a pretty good season for gaming graphics cards, and GTX 460 is really the best graphics card we've ever built for gamers in the history of our Company.

So that's one.

The second one is Quadro Fermi.

Fermi-based architecture Quadro is just ramping into production now, and the OEMs are really enthusiastic about it.

The end users are clamoring for it, and so I think this is an opportunity for us to upgrade the installed base of Quadro workstations around the world with a revolutionary new architecture.

Mike already talked about in his comments pretty substantial speedups from generation to generation.

I don't think we've ever seen the 5-X speed-up from one generation to another generation, so this is a very, very big deal.

And then the third thing is Tesla.

We had a record quarter in Tesla this last quarter.

We've got some exciting new products that we're going to ship into the marketplace imminently.

And so I think this is a marketplace that is clamoring for more processing capability, and we are not seeing any effect from the consumer marketplace because of it.

And then the last part is Tegra.

As you know, the most important and probably the fastest-growing computer market today is the smart phone and the tablet market.

And Tegra is our entry into that.

We are seeing customers trying to pry these chips out of our hands, and we've got to get these things shipped into production.

And they are really exciting new products.

So those are the four areas where we are confident to see growth -- GTX 460, Quadro, Fermi-based Quadro, Fermi-based Tesla and our Tegra SOC.

With respect to gross margins, all of those products that I just mentioned are either at or above the corporate gross margins that David talked about.

Dave, if I could just follow -- I appreciate that -- just follow up on the consumer PC, the GTX 460 really driving a lot of this growth -- given the uncertainty in Europe and China, and given the fact that you are talking about attach rates of discrete graphics potentially coming down or came down this quarter on the exchange rate, what makes you confident that you can get the attach rates up or improved demand in those two key segments for you?

Well, there's really two GeForce markets.

One GeForce market is the gamers, is the GeForce GTX market.

They are not sold to OEMs.

They are not really sold to system builders.

They are sold to end users.

Most of it is sold through e-tail or retail.

The vast majority of it is sold as a retail-based add-in card.

In a way, the way you can think about GeForce GTX is that it's the game console within the PC.

What really drives the sales of that segment of the marketplace is games.

There's another part of our GeForce business, the lower-end GeForce business, that is sold to OEMs.

They way that the OEMs think about GeForce in those markets is that it's the ultimate, if you will, or the premium -- the ultimate upgrade or the premium experience PCs.

Every PC is built from either AMD or Intel CPUs and various integrated graphics.

And if you wanted to create a product line with basic and competitively priced PC on bottom and premium experience PCs on top, the best way to do that is put a high-resolution display on it and then maybe you add 3-D Blu-ray to it and you add a discrete GPU to it.

And that's really how the vast majority of the PC OEMs, as you look around the world, differentiate their premium PCs from the commodity or baseline PCs.

And so, if you look at that marketplace, that's what I think is uncertain and what -- we will continue to assume that it's uncertain until it's not.

We saw it probably -- we saw it before everybody else did.

And the reason for that is because we are in the high end consumer PC segment, and as memory prices went up and as the euro got week and the various end markets got soft, the first thing that people did was shifted down price and into integrated graphics.

And then, of course, if the weakness persists, eventually the total number of units of PCs will decline, which is what people are starting to see now.

So I think we're taking a relatively conservative posture with respect to the PC OEM business, but the four businesses that I talked about just now, I think, are businesses that we feel pretty good about.

Hans Mosesmann, Raymond James.

A couple of questions -- can you comment on how you see PC seasonality generally, and also can you comment on the prospects of you doing a buyback?

The second question I'll just take very quickly.

We talk about these topics at our board meeting, every board meeting, and so the use of cash is a topic that comes up on a regular basis.

And so we'll have that conversation again at the next board meeting and do the appropriate thing.

With respect to seasonality, for the overall consumer PC market we are taking a rather conservative posture about it and we are assuming that what we saw before everybody else, this is a -- let's see.

We started seeing weaknesses in the end market probably starting in May, and May-June time frame, late May/early June time frame, and that has persisted since, and I think other people are starting to see it across the board now.

We're going to take a relatively conservative posture with respect to that segment of the market.

The four businesses that we see growth in we feel pretty good about, and they are not affected by the end consumer market.

The vast majority of that is enterprise business, and although the consumer GeForce business or the GeForce GTX business has targeted gamers, what really drives the sales of that is great games.

And it's hard to find a better game than Star Craft II, to propel the upgrade of graphics cards for gamers.

And then the last one, smart phone growth and tablet growth -- that's the most important segment of the computer industry today and it's the segment that we are entering for the first time, starting this last quarter in a small way and then with this quarter hopefully in a much larger way.

Daniel Berenbaum, Auriga USA.

Just to clarify, you talk about gross margin would have been above expectations if not for the inventory charge.

What was the inventory charge?

[Hans], as any company in our business, we have inventory write-offs every single quarter.

And so the question then becomes, what's extraordinary.

And rather than be subjected to how we determine that, I think the easiest thing for us to do is just simply say that, no matter how you might calculate what the exceptional piece was, it was large.

And excluding it, we would have exceeded our guidance.

And as far as the specific amount, that all depends on how you would want to make the calculation.

Okay, so just on my rough calculation, if I back you out to get to above 47% gross margin, that gets me to the place where you had a $75 million inventory charge and your days of inventory would have been around 110 days.

Is that the right order of magnitude and the right way to think about it?

It's in the range.

Okay, and then just to follow up, you had commented that we've got these new games, Star Craft coming out.

How do you think about the advent of programs like OnLive, which centralizes all of the compute and graphics functions and where I'm told that you don't need a strong GPU at the local desktop or you can stream HD graphics straight to the desktop without actually having a graphics card on your PC?

How does that affect how you think about games going forward?

Well, the way to think about OnLive is, imagine it's like subscription TV, and many people have subscription TV as well as premium content.

The way that OnLive works is fabulous, and we are wonderful partners with them.

What you see in the servers is a bunch of GeForces.

And so we are excited about the work that they are doing, we are excited about getting video games to a larger number of people and making video games much more accessible.

So all in all, I think it's a fabulous development to have cloud-based GPU computing, if you will, emerge.

But the services still needs to reach a much broader audience.

Latency matters, and so as OnLive builds out their servers and server farms around the world over the course of the next several years or a decade, people still need to be able to enjoy games wherever they are.

And so I think that's one factor.

The other factor is not all games will be available from OnLive.

Star Craft II, for example, is not, and WOW will likely not be.

So if you want to enjoy a broad range of games, having the ability to play the game on your local PC as well as being able to serve it off the cloud is a wonderful, wonderful way to enjoy more games.

So we are a huge proponent of it, and we work closely with them to make the service as good as possible.

And in the short-term what we see is a growth opportunity in putting GPUs into the cloud.

Ross Seymore, Deutsche Bank Securities.

On the inventory charge side of things in the gross margin, is the gross margin benefiting at all in the third quarter because of written-off inventory?

No.

We wrote off what we anticipated would not be sellable, right?

And so it would be inconsistent to assume in one quarter that you're going to write it off, the next quarter you're going to sell it.

So we wrote off what we believed was in excess, so our guidance doesn't include any of that selling through.

Again, I heard the earlier question about you not really want to sizing things precisely, but could you give us an idea of what the average charge for inventory or write down would have been over the last few quarters just so we can get a ballpark estimate about what it was this time?

No, I don't think we've disclosed that.

Maybe last question, then, a quick one -- any help on the graphics side versus the chipset side?

Any difference in trends in the quarter there, sizing the remaining chipset business you have?

The vast majority -- our chipset business was okay, and I think its dynamic is a little bit different than the overall discrete consumer PC GPUs.

Our chipset business is predominantly chipsets that go into Apple computers, so that's one way to see its progress.

Rick Schafer, Oppenheimer & Co.

This is [Sean Simmons] calling in for Rick.

I just had a couple of questions here.

What do you guys see the impact from the Intel and FTC settlement recently announced?

Is that going to help you with your ongoing trial with Intel?

The two different cases are independent and not the same.

The first case is obviously the US government finding Intel to have practiced businesses and conducted itself in an improper way.

And so whatever it is that the government does to enforce and whatever remedies that they've suggested to Intel or compelled Intel to follow are all good things.

To the extent that Intel becomes a much better citizen in the ecosystem and conducts business in a proper way, that's good for the entire ecosystem.

It's good for us, it's good for AMD, it's good for customers, OEMs.

It's ultimately good for consumers.

Our case with Intel is not related to that, and it's strictly related to a contract dispute related to, on the one hand, chipset license to us and, on the other hand, a cross-license to them.

And so that particular dispute goes to court, goes to trial, I guess, towards the end of the year.

So we're looking forward to that.

Okay, great.

And then, I guess looking at your just core GPU business, you've been talking a lot about the consumer side and then the gamer side.

Is there any way we can figure out how much of a percentage of your sales comes from, I guess, the gamer side and how much you have for the PC OEM side?

I would say that it fluctuates from time to time.

Obviously, Fermi was late and GTX 470 and GTX 480 was about six months late, later than we liked.

And so we lost share in the Q1 time frame.

We started gaining share back in Q2 and we're going to gain a lot more share back this quarter with our GeForce GTX business, which is targeted at gamers.

And so it's hard to say based on the last several quarters what the percentage is.

What we do know is that we're likely grow into that marketplace and to continue to take share for the remainder of the year, as far as we can see right now.

Krishna Shankar, ThinkEquity.

Going forward, what do you think is the attach rate of discrete GPUs now in notebooks and desktops, and how do you see that trending over the next 12 months or so, especially with the coming Sandy Bridge and Fusion platforms targeted at low- to mid-range PCs from both Intel and AMD?

The attach rate of discrete GPUs this last quarter I believe actually went up.

The number of discrete GPUs in the marketplace was about the same.

And so, for the last several quarters, I don't believe -- I was shown some Mercury numbers which unfortunately doesn't account for our fiscal quarter versus AMD's calendar quarter.

But anyways, the basic math here is still relatively the same, that the GPU discrete marketplace is -- the number of discrete GPUs totally shipped is about the same from the last several quarters, and our expectation remains the same in the next several quarters.

Notebooks is way up.

Discrete GPUs for channel desktops was down.

We believe that a lot of that has to do with the fact that DRAM prices had gone up, the euro had weakened and so people had to skew down.

And so, looking forward, my expectation is that OEMs are going to continue to want to segment their PCs.

And for the mainstream segment, where the basic performance is what comes with integrated graphics, they will continue to use integrated graphics.

And for the premium segments, for the premium experience PCs, they will want to have a discrete GPU to both differentiate that platform as well as deliver a better experience.

My expectation is that that desire by the OEMs to differentiate, to segment, to offer a premium experience is not likely to go away.

SandyBridge is going to increase integrated graphics performance just as every generation that have increased integrated graphics performance.

So by the time that SandyBridge ships there's going to be some pretty fabulous discrete GPUs that ranges from the basic levels that we offer today to the high end.

So we will see how it turns out, but my expectation is that the OEMs will continue to want to segment and differentiate their platforms.

My follow-up question is on Tegra.

Can you give us some sense for the higher-volume production here in the October quarter?

Is this more smart phones, Web tablets?

Can you give us some color as to the kinds of OEMs you are engaged with?

Yes, just like you said.

So both smart phones and Web tablets?

We hope so.

They won't be of equal proportion right off the bat, but we will see which devices ship first.

But whatever devices ship, they are going to be pretty amazing.

Alex Gauna, JMP Securities.

I believe you mentioned your confidence that you can take share back in the back half of the year, but you admittedly said that your products are skewed towards the high end.

I wouldn't argue that the 460 is going to be a category killer.

Can you really take share broadly with your product lineup the way it is, or should we expect that we're going to see the Fermi get into some of the more mainstream SKUs?

We are ramping Fermi from top to bottom on desktop, on notebook, on workstations.

With respect to market share, we probably lost a couple, 2 million-3 million units, if you will, in market share.

And that's a lot, that's a lot when you think about it that way.

The vast majority, as you mentioned, of all of that is low-end discrete.

But the price of those GPUs are low teens, anywhere from $10 to call it $12-$14.

And the gross profit dollars on those devices are not extremely high.

And so we know how to go win that business back.

During the first quarter or so, because we were so constrained in supply and the market was so robust, we were reluctant to be too aggressive in capturing business that we wouldn't later be able to fill.

As it turns out, as the markets softened in the May time frame and the June time frame, that turned out to have been a bad decision.

And so we are in pretty good posture now, and we know how to go take share back in the low end.

So we'll take some share back in the low end, but more importantly the profitable segments and the high ASP segments -- we just have a great product line.

So that's our basic approach, is to offer the best products in the segments that are targeted to gamers.

And for the segments that are targeted at OEMs for their premium upgrades and premium experience PCs, we have very price competitive products that we can go take share back with.

In that OEM category, does that start by Q4, or does that wait more until Sandy Bridge in 2011?

It starts right away.

I think part of our business is still --

Thanks very much.

Yes.

Craig Berger, Friedman Billings and Ramsey.

Can you help us understand how big your chipset business is at this point and how that should trail off over the next, say, year or two?

Let's see.

It's less than $200 million a quarter and we will remain at that rate for probably through Q1 or maybe even Q2 of next year.

And then it will start trailing off not completely to zero because we will continue to have AMD integrated graphics solutions in the marketplace, and our expectation is that we're going to continue to sell those for probably through the end of next year.

So that's my expectation of it.

It will start declining call it Q2-ish and then decline towards -- to something a lot less than that by the end of the year.

As a follow-up, the workstation business has ramped pretty meaningfully.

Is this a sort of run rate sustainable level of business here?

Is there some catch-up that's benefiting the numbers right now?

And then just one last follow-up after that.

Our workstation business has been growing consistently quarter to quarter, but it hasn't reached the record levels that it was at a couple years ago.

And so my expectation is that the workstation business is going to continue to grow.

It's going to continue to grow for two reasons.

One, more and more people are upgrading to Windows 7, and that upgrade process will affect workstations as well.

The second reason is during this time, this last couple of years, we have invested in expanding the reach of our workstation solutions.

Whereas it was predominantly used for manufacturing in the past, our workstations are now used for broadcast workstations, for 3-D television streaming, for post-production, for video editing, medical imaging -- gosh, the computational simulations of fluid dynamics.

There's all kinds of things that our workstations are now being used for.

And so, on the one hand, I think that the upgrade cycle is -- we are benefiting from the upgrade cycle.

And then, secondarily, we have invested in expanding the reach of our workstations solution.

So we're just exposed to a much larger market.

And so my expectation is that our workstation business will surely recover to its record highs and go well beyond that.

The last question is, you guys referenced macro weakness in May and June.

Your largest competitor did not have similar commentary, and I'm wondering if there's other dynamics at play or what might drive a difference between you guys.

I think that -- well, first of all, I don't completely know.

We're going to see all of this stuff sort out as we see their Q3s.

You have to remember that our business is skewed off by a month, and being skewed off by a month in this last quarter is a pretty big deal, as it turns out, because the market softened in the second month of our quarter, which was practically the end of theirs.

And so I think we are going to find out how it really sorted out probably by Q3 time frame.

But we saw market weaknesses, not just for ourselves in the end markets, we saw them for other people.

And we were, frankly, surprised by their guidances.

But so be it; we'll find out when everything sorts out.

Do you think the weakness is driven by demand, or is it supply chain adjustments?

It's end market weakness.

I think at this point, I think most of us believe that Europe in Q2 was soft, that the euro was weak.

And that naturally would cause some skew downward.

And China, reducing their stimulus to the economy have slowed down the market a bit.

Now, of course, our business is -- we have a much larger exposure in Europe and China because our brand is so useful and so powerful to help the local OEMs enhance their brand, and also because consumers are much more, much more -- find video games and content much more appealing out in the other markets.

I think that those are just incredibly important markets to us.

And so, our exposure is probably larger than theirs.

Overall, I think that we're just going to have to wait and see how their Q3s shake out for us to really know.

Glen Yeung, Citigroup.

Jen-Hsun, can you discuss whether or not in your past experience in periods where the economy seems to be decelerating, whether gamers are resilient to that, or do they have to follow the economic condition?

Well, I think that there is some impact, surely, to the economy, but there's nothing like great games.

And to that extent that -- it's kind of hard to think about Star Craft II as a game.

I really think of Star Craft II as an event.

This is a once-a-decade refresh to one of the most important video games of all time.

Star Craft, as you know, is a cultural phenomenon in many countries.

They watch Star Craft competition on TV.

There's a Star Craft channel.

There's Star Craft teams.

So in many countries, Star Craft is just a very, very big deal.

And so this is very welcome upgrade, and the game is absolutely beautiful, and GTX 460 is perfect for it.

So my sense is that for the 12 million Star Craft players out there, there's some serious upgrading to do.

And my sense is that that should far overcome any softness in the marketplace.

You also sound quite confident that Tegra will grow in the quarter.

And I wonder if that prompts you to adjust your Tegra run rate forecast.

I'm not sure I understand the question.

Glen, I think we've given guidance on that in the past, and I think at this point it would be early to call whether we are going to exceed that or not.

So I think we hold to that.

We still feel very bullish about the product, obviously.

Arnab Chanda, ROTH Capital Partners.

I have two questions.

One is, Jen-Hsun, I know you don't want to give specific guidance, but could you talk a little bit about by the time the chipset business either becomes really small or maybe even goes away, do you think Tegra and Tesla would be big enough to offset that?

And I have a follow-up, please.

My sense is that the chipset business will be about half to a quarter its size in Q4 of next year.

And so, if you roughly said, it was about a quarter of the size, would Tegra and Tesla be able to make up for the difference?

Both in revenues in surely much, much more in terms of gross profit dollars, my sense is yes.

Okay, great.

And then one other question about your Quadro and Tesla business -- do you think that -- is Quadro -- are you at a point where you have basically kind of completed the penetration in the market and it's really more about Tesla for that entire product line?

Or do you think there's multiple opportunities left in terms of growth in that business?

Quadro has wonderful growth opportunities ahead of it.

First of all, where -- people buy workstations and upgrade their workstations on the rhythm of the companies themselves.

The automobile industry that uses CATIA or the airplane industry or the tennis shoes industry -- they all upgrade on different cycles, and they use different packages.

And more and more industries are using digital approaches to prototype or style or simulate, just like we do.

In your industry you use computational finance.

In my industry we use EDA simulations and EMI simulations.

We use all kinds of simulations to predict the outcome of our product.

And whether it's in design or manufacturing or advertising or in broadcast or in medicine, more and more people are using digital approaches to predict the outcome of the products.

And so I think that the market is surely expanding.

The other factor that's really exciting for us and what's so great about Fermi Quadro is that it's such a great discontinuity relative to anything that we've shipped before.

I don't know that we've ever shipped a product that is five times faster in the type of applications that it's used in, from generation to generation.

So we are looking forward to upgrading the installed base in addition to the natural growth and the natural upgrade of the workstation market.

So I'm really excited about Quadro, and we're looking forward to seeing how it does in the marketplace.

The OEMs are really excited about it as well.

We have HP and Dell and Lenovo all geared up, Fujitsu Siemens all geared up to -- these are our workstation partners -- geared up to take Quadro Fermi to the market.

James Schneider, Goldman Sachs.

I guess just to start off on the notebook GPU business for a second, can you give us a sense, if I just isolate that -- you mentioned that the market itself for notebooks is up in the quarter.

Was your business also up for notebooks in the quarter?

And then, do you expect the notebook business in Q3 to be better or worse than your guidance for the corporate in Q3?

What was the last part of the question?

James, are you talking about the Q3 guidance?

Yes, your Q3 guidance -- do you expect notebooks to be up in Q3 or essentially better than your corporate guidance for Q3?

So if you look at Q2, notebooks and desktop were both down.

Obviously, we missed our original guidance by a fairly large amount, and both of those were key contributors to that.

If you look at our guidance into Q3, we are expecting both of them to be up -- obviously not by a lot, given that we are only guiding 3% to 5%.

And then it looks like you're OpEx ticked down a little bit in Q2.

Can you give us a sense of what expenses you may have cut back on in Q2 and then how we would expect -- what marketers you look for as business recovers, how you bring back spending incrementally and what are the things you bring back first versus later?

Well, if you look at the expenses that we've worked on reducing and so forth, they cover a broad range of things.

They cover things in terms of how we take out products, how we kit prototype for engineering test vehicles.

It covers the minutiae things, like travel and the things that are not very exciting.

But we've been fairly diligent across the entire business trying to figure out how do we reduce waste, how do we economize in terms of our expenses without, naturally, impacting our development plans and so forth.

And going forward, right now certainly in the environment we are in today, we don't see our attention on expenses abating anytime soon.

We do have continued efforts in reducing cost, but we also have cost pressures going in the other direction.

And so our guidance is a recognition of the fact that we think, for the time being at least, that the cost reductions that we will implement in the third quarter would offset some of the upward cost increases going in the other direction.

Shawn Webster, Macquarie.

Within the cost of goods sold for that $181 million, how much of that was due to be product recall charge?

You mean that versus the litigation?

Is that what you mean?

Is there litigation within the cost of goods sold as well?

Yes.

Yes.

How much were those two combined?

We don't really break that out, but let me be a little bit clearer.

When we talk about litigation, there's the cost of the litigation itself, but then there's also the cost of the remedies.

Right?

And so the remedies have to do with repair costs and so forth.

And so there's a piece of it associated with cost of goods that's naturally as a result of the settlement itself.

And we don't break that out.

Oh, okay.

And for your inventories going into Q3, do you expect them to be flat, down or up?

Probably a modest change, only at this point.

A modest change up?

When I say modest, it will be either modestly up or modestly down; it's within a -- I would basically say it's basically flat.

Okay, got you.

And I understand there is a mix effect happening in the quarter.

Can you tell us what your GPU ex-the MCP business, the pricing, did sequentially for Q2?

No, we don't disclose that.

Okay, how about tax rate?

So the tax rate is coming up.

Can you share with us why that's popping up?

Yes.

So if you look at our business, our taxes are basically a function of how much income we make.

And as our income goes down, some of the jurisdictions where we operate and pay taxes in -- that income doesn't necessarily go down.

So the taxes there, you might almost view as a fixed cost to the Company.

So as our total overall income goes down, those taxes stay flat, and hence are rate goes up -- and vice versa.

When our income goes up, the rate goes down.

Kevin Cassidy, Thomas Weisel Partners.

Just two quick questions -- on enterprise, last night, we heard from Cisco saying that they're seeing a slowdown and enterprise.

Can you describe what's different?

Maybe it's the product refreshes causing more demand for your products.

Or, maybe if you could describe what's happening geographically?

Well, our enterprise business is a little different than an enterprise desktop or a router, per se.

If you look at our workstation business, the workstations are used to create the essential products of those companies.

And to that extent that those companies believe that their workstations are no longer high-performance enough or capable enough to handle the level of complexity or the amount of simulation they would like to do in simulation, they will naturally upgrade.

Frankly, most of these workstations these days -- by buying the workstations, you are reducing the cost of development of most of your projects because you are just able to do more in the digital world where virtual prototyping is just much more cost effective than real prototyping.

And so it's not a clear comparison to us, and right now we are not seeing changes in the nature of the business and the various markets.

We are frankly too busy right now ramping Quadro based on our Fermi architecture, and so the enthusiasm is quite high right now.

And maybe if there -- do you see any new competition or any changes in the landscape there?

Well, the workstation market is not about a graphics chip and is not about a graphics card, and this is something that the vast majority of our competition just no longer understands.

People aren't buying our Quadro solution a chip at a time.

They're trying to solve a problem, and it could be that they are trying to ray-trace a car so that they could see a photorealistic image of it while they are trading off various designs.

It could be that the data set is just absolutely enormous and they need to have the scalability and the scalable solutions that we provide.

It could be that somebody's workstations needs to be virtualized and have a Windows NT on one or a Windows or Linux on the other, and both of them has to be accelerated.

And the fact that we have the best software stack for both Linux as well as Windows in the world makes a huge difference.

And so the list of reasons why people choose Quadro is quite deep.

This is an area where we have a great deal of passion for and one in which we do an extremely good job.

And so the bar is pretty high, and we are not seeing any significant competitive changes out there that we notice right now.

Patrick Wang, Wedbush Morgan.

Jen-Hsun, can you talk about your -- I guess just a high-level view of the competitive landscape in graphics -- I guess what is your best sense in terms of the pricing trends?

And I guess what kind of pricing you're baking into your gross margin guidance?

What kind of pricing -- well, we have more high-end exposure today than we've had in a couple of quarters, and so that's a positive thing.

On the mainstream product, if we want to go gain share back, we'll sell several million more lower ASP products, and so that will have a negative trend on ASPs.

And so we'll see how they all add up.

But we are relatively comfortable with the gross margin guidance that David gave earlier.

Yes; the guidance contemplates that.

I see, got you, that's helpful.

And then just for my follow-up, I guess in your prepared comments you stated that channel inventory levels had come down.

Can you help us understand how much the product mix of that inventory and how that compares to, I guess, levels that are comfortable or maybe historical for you guys?

So we get inventory reports out of the channel on a fairly regular basis.

So I can't tell you, other than from an aggregate standpoint, that generally the trend has been down for the last three months.

We expect it to go down again in the third quarter.

But, even though it is trending down, it's relatively at levels that would be consistent with historical averages in the industry over time, which have been traditionally somewhere around in the seven-eight-week type of level.

We've certainly had periods when it's been a lot lower than that, we've had periods when it's been well above that.

But today, it's relatively close to the historical averages.

Patrick, if you think about channel inventory, because most of that inventory is traveling by sea for the lowest possible cost freight, some three to four weeks of that seven to eight is material that is sitting in the ocean going across.

Right?

And so, what is really available in the channel is about three to four weeks or so.

And so I think, if we are sitting at these levels, we all feel pretty good about it.

There are no further questions at this time.

I will now turn the call back over to you, sir.

Thanks, everyone.

We look forward to talking to you about our next quarter results.