Highlights From NVDA's Q3 Conference Call: Steady Increase in Demand Driven By Healthy Marketing Environment and New Apps

November 6, 2009 12:54 PM EST

NVIDIA Corporation (NASDAQ: NVDA) reports Q3 adjusted EPS of $0.19, 9 cents better than the analyst estimate of $0.10. Revenue for the quarter was $903.2 million, which compares to the estimate of $835.19 million. Shares are up 7% today

Highlights From NVDA's Q3 Conference Call:

• GP demand continued to steadily increase, driven by a healthy marketing environment and new applications that require GPUs. Our unprecedented investments in parallel computing are creating change and opportunity, and are the foundation of our growth in the coming year.
  
• We believe we are in the midst of a giant leap in computer graphics. The GPU continues to advance the most parallel of applications, 3-D graphics. And it is now poised to revolutionize a wide range of industries by making parallel computing mainstream.
  
• Movie production houses that worked on Terminator Salvation and Star Trek ran color grading and image processing 20 times faster using parallel computing on NVIDIA GPUs.
  
• We held our first GPU technology conference last month, and even after closing registration early, we were oversubscribed by 50%. 1,500 people from over 40 countries listened, learned, and shared during 200 hours of sessions. At one point, more than 35,000 people were watching online.
  
• (CFO) Revenue was $903.2 million, above our guidance and up 16% sequentially, making the third consecutive quarter of top line strong growth.
  
• Gross margin was 43.4%, 41% on a non-GAAP basis, significantly higher than guidance.
  
• GAAP operating expense was just under $284 million, marginally above guidance. And GAAP net income was $107.6 million or $0.19 per diluted share. Non-GAAP income was also $0.19 per diluted share.
  
• Our core GPU business was particularly strong, up almost 25% sequentially. Within that, our desktop and notebook GPU segments were up 19% and over 41% respectively quarter-over-quarter. While demand was strong, we were supply-constrained throughout the quarter, particularly in our 40-nanometer products. 40-nanometer products for desktop and notebook constituted 19% of our total GPU revenue in the quarter.
  
• Our professional business saw another quarter of growth, up 11% sequentially. While revenues in this business are still well below their levels of a year ago, quotation activity continues to gain momentum and is encouraging.
  
• Gross margin improved sequentially as a result of several factors: significantly higher than anticipated unit sales of GPUs; improved 55-nanometer yields, as well as other cost reductions; and more favorable product mix across businesses, and particularly better mix within the GPU business itself.
  
• Inventories at the end of the quarter were essentially flat compared with the prior quarter.
  
• Inventory days on hand improved to 47, calculated on a non-GAAP basis, a four-day improvement over Q2. Inventory in the channel remains very low at around a month.
  
• With demand outstripping supply, we don't see OEMs at the channel building inventory.
  
• Our outlook for Q4 of fiscal 2010 is as follows. Revenue is expected to be up slightly, approximately 2% from the third quarter. While revenue from 40-nanometer products is expected to grow significantly quarter-over-quarter, we expect 40-nanometer products to be supply-constrained throughout the quarter.
  
• GAAP gross margin is expected to be 40 to 42%, essentially flat with Q3, excluding the insurance reimbursement. GAAP operating expenses are expected to be in the range of 305 million. This is up from the third quarter, due primarily to our fourth quarter comprising 14 weeks versus the usual 13
  weeks, as well as costs relating to a record number of product bring-ups.
  
• (Q&A) I guess I would start with - we've seen several quarters now where you've significantly exceeded your guidance. Can you talk a little bit about the business processes that you have in place to forecast your business and whether you're finding yourself sort of short of product going forward as a result of not really getting accurate forecasts from your customers in your channel? (A)So, Doug, our process starts really at the very baseline of our field sales force, and we gather forecasts from the field, both the OEMs and to the extent we can through the channels. We also at the same time try and correlate that with what market forecasters expect to be growth rates in the various end markets we participate in. And from that, we set revenue plans. And from that, we set wafer start plans ultimately. I think the challenge for probably everyone in the industry right now is that demand has been more robust than what anyone has been able to particularly foresee. And given that you've got a 12-to-13-week lead time from the date you decide that you want more volume to the date that you can ship that volume, we're being surprised by robust demand within that lead time. And so it's a good thing in some respects. I think the downside for us has been that, as a result, we have been supply-constrained. But we do look at a number of different inputs as it relates - in terms of compiling our forecasts and trying to challenge our crystal ball, so to speak.
  
• All right, if we could sort of move in and sort of focus in on one segment of the business, the chipsets, it appears during the quarter you guys have made a little bit of a change in tone towards the outlook for your chipset business. Can you talk about that? And can you talk a little bit about what's going on over with the AMD chipsets, the downturn there you attribute to their supply constraints? Is there any thought that you might be also possibly experiencing some share loss in that piece of business? (A)Our primary AMD chipset today is a chipset called the MCP61. The MCP61 is the only single-chip solution in the marketplace today, and it enables the lowest cost solution for an AMD platform. There's really at this point nothing that I know how to do, nor do I believe anybody else can build a chipset that is as low cost at a system level as MCP61. You know that the AMD market is primarily mainstream desktop. And then you also know that for a variety of reasons AMD CPUs lost share last quarter. And so when AMD CPU loses share, they lose share primarily in the mainstream market, and so that's basically what we experienced. I don't think we lost any share within the AMD CPU market. I think the AMD CPU lost share in the mainstream desktop PC market. If the AMD CPU recovers in share, relative to Intel, we'll grow the MCP61 business again. And so my expectation is that for the foreseeable future, certainly well beyond several quarters, the MCP61 is going to be a wonderful chipset for AMD processors. The second question is longer term about our chipset. You know that we have a dispute with Intel. And it is now impossible for us to build next-generation chipsets with all the allegations that they're making in the market and with the dispute hanging over it. We have one major generation of the current architecture post MCP79. It is a really exciting new MCP next-generation, and we're expecting it to be very successful. The customer feedback has been quite wonderful.And while MCP79's GPU capability already far exceeds that of Arrandale, MCP89 will take it that much further. And so MCP89 is our next major generation of this architecture. And so we're expecting our chipset business overall to do quite nicely through next year and beyond. The only thing that I have said, and I want it to be very clear, is that we have no intentions at this point of building a DMI-based chipset. But otherwise we have ongoing investments, as well as chipset developments in the current architecture.
  
• I guess from an end demand perspective, first and foremost, beyond the limitations on the supply side, are you seeing any sort of abnormal seasonality or inventory-related impact on demand as we look into the fourth quarter? (A)We came into this quarter with very, very strong demand again. Last quarter, we were supply-constrained from the very first day of the quarter, and it got more and more challenging as demand continued to grow. I think that Mercury Research showed that we lost market share. I think we did lose market share. We came out of the quarter not being able to fulfill several million units of demand on us, not to mention the terms business that would've been on top of that. And so the demand on our GPUs at the moment is very, very, very, very high. Is it because of Windows 7? It could be. I hope that it is. And I hope that it's because people's PCs are getting rather old, and they could use a refresh. I hope that it's because the adoption of GPUs is higher than before. And so we're seeing very, very strong demand at the moment. I expect us to be supply-constrained, and the projections that we're giving you guys are the supply-constrained forecasts. And so I'm expecting us to be supply-constrained, and we're going to fight really hard to make up as much as we can. But the demand is pretty strong out there right now.
  
• I guess as a follow-up, also looking at your MCP business, it was helpful to hear how you think it's going to play out. But if we try to get a little more granular on that, the fact that you're not planning to do any of the DMI side of things - you're at 250 million in revenues now. When do you think it starts to impact your overall chipset business that you're making the choice, or Intel's forcing the choice, that you won't be supporting the new architecture? When would we start to see that on the revenue side of the equation? (A)My sense is that we'll see our - and I'm just giving you my crystal ball. Who knows? But based on our analysis, my sense is that our chipset business ought to peak probably around Q3, Q4 next year. And then they'll ramp-down from there. And it just depends on available CPUs that we can connect to, customer demand for Ion, so on so forth, our competitiveness relative to Arrandale, Arrandale CPUs. Arrandale has integrated graphics, but MCP79 already has better GPU technology, and MCP89 takes that even further. And because Windows 7 requires direct compute and open CL and direct compute are becoming increasingly important, what we call GPU computing, I think that the disparity between our MCP and Intel's integrated graphics is growing, not shrinking. And it's growing by leaps and bounds at the moment. And so I think those are all play factors. But without much estimation, I think that our chipset business is likely to continue to grow for the foreseeable future, or has the opportunity to grow certainly.

Related Categories

Corporate News
Earnings

Stocks Mentioned

Add Your Comment

Name

Subject

Body

Verification can't see the text?
click here to refresh the image.
what's this?