conference date: August 6, 2009 @ 2:00 PM Pacific Time
for quarter ending: July 26, 2009 (2nd quarter fiscal 2010)
But I own a competitor, AMD.
Overview: Pretty good recovery sequentially, but still below year-earlier levels and still losing money.
Basic data (GAAP) :
Revenues were $776.5 million, up sequentially 17% from $664.2 million in Q1, but down 13% from $892.7 million in the year-earlier quarter.
Net income was negative $105.3 million, up sequentially from negative $201.3 million, and also up from negative $120.9 million year-earlier.
EPS (earnings per share) were negative $0.19, up sequentially from negative $0.37 and also up from negative $0.22 year-earlier.
Revenue in fiscal Q3 2010 will be up sequentially 5 to 7%. Gross margins will be up.
They say business is recovering and demand improving. New products are now generating significant revenue: Tesla for high performance computing and Tegra for the mobile market.
A charge of $119.1 million was made releated to the past problem with weak die packaging materials. "The cost of remediation and repair of impacted systems has been higher than originally anticipated."
Cash and equivalents ended at $1.47 billion. Free cash flow was $117.5 million.
First 40nm GPUs were shipped in the quarter. Quadro FX SVS scalable visual computing platform was launched. Tesla solutions are now carried by HP, Supermicro, Cray, Dell, HP, Lenovo, SGI and Sun.
Cost of revenue was $619.8 million, leaving gross profit of $156.7 million. Operating expenses of $266.8 million included $192.9 million for R&D and $74.0 for sales, general and administrative expense. Operating loss was $110.1 million. Interest income $3 million. Income tax benefit was $2 million.
36.3% non-GAAP gross margin. Non-GAAP net income $37.7 million; EPS $0.07.
GPU computing has surely reached a tipping point. Almost every video acceleration program now uses CUDA; many other applications as well in science and finance. Windows 7 (Direct Compute) and Apple Leopard will be ready for CUDA. Integrated graphics chips will not support Direct Compute.
Tegra has been chosen for many next generation cell phones and portable devices. 26 ION platform designs have been announced to date.
Quadro revenue has declined more than 40%, but was flat sequentially.
Tesla had about $10 million in sales, its first significant quarter. We believe this is a $5 billion market opportunity over the next 3 years.
GPU business was 48% of total revenue. Up 5% sequentially. Desktop up 5% sequentially. Channel inventories remain at about 1 month. Notebook revenues fell 5%. Lost some share. Fall platform 40nm GPUs should regain share
MCP business 31% of total revenue, up 27% sequentially. Chipset products for AMD desktop processors were up 55%, but the Intel segment also grew.
Professional business was 15%, up 10% sequentially, almost the increase was due to Tesla. Quadro demand bottomed.
Consumer segment was 6% of total revenue.
$119 million charge? We made an estimate for this problem a year ago, when we had little data. We have now made agreements with almost all of the affected OEMs, so our estimate shoudl be far more accurate. Bump crack issue is not very distracting for the company; we are focussed on the future. It has not affected our position in the marketplace.
Market share versus prices on GPUs? We moved more chips into 55nm node. Tegra and Tesla have better gross margins, so as they ramp margins should improve. A couple of quarters ago we had a lot of 65nm inventory, which we have largely worked through. We should be able to continue to have a premium price and still capture market share.
3D vision capabilities to be included in GeForce will be a watershed event later this year. We are miles and miles ahead of our competition.
Attach rates for discrete GPUs in recessionary environment? We have not noted a shift. We know we have better attach rates in some segments than other.
40nm progress? We are by far the largest producers of 40nm today. By q4 it will represent a large portion of our revenue, but not a majority. Margins will be higher.
Tesla servers consume far less power than conventional servers because of their performance capabilities. Every company that designs products is going to want our new technologies.
Dispute with Intel about chip sets? We are not commenting about the future. We are not necessarily supporting Intel with chipsets in the future. We are very clever and you will have to wait for the punch line.
We only do well on AMD chip sets when AMD's low and mid-range CPUs do well. So we are rooting for them and would like them to do well. Our aim is to enhance baseline capabilities of the chip sets over what AMD or Intel provides.
Why not guide to higher sequential growth when Q3 is usually your strong seasonal quarter? We don't have any particular deep insight into the economy. Historical seasonality is not a guide this year. We guide to the best of our ability.
NVIDIA Analyst Conference Summaries Main Page
NVIDIA Investor Relations page
Openicon NVDA main page