Analyst Conference Summary
conference date: October 15 , 2009 @ 2:00 PM Pacific Time
Overview: Sequential improvement in revenues, but still down from year-earlier.
Basic data (GAAP):
Revenue was $1.40 billion, up 18% sequentially from $1.18 billion, but down 22% from $1.80 billion in the year-earlier quarter.
Net loss was $128 million, improved sequentially from a loss of $330 million, and also improved from a loss of $134 million year-earlier.
EPS (earnings per share) were negative $0.18, improved sequentially from negative $0.49, and also improved from negative $0.22 year-earlier.
"AMD expects its Product Company revenue to be up modestly for the fourth quarter." $500 million in operating expense. $100 million capital expenses for full 2009.
It was a profitable quarter if you focus on the Product Company, due to strong product demand and "disciplined execution." "Improved factory utilization rates, higher microprocessor average selling price (ASP) and an increase in 45 nm product shipments resulted in a gross margin improvement." Gross margin was 42%, up from 37% sequentially.
GAAP results include the foundry segment; non-GAAP is limited to the Product Company.
Consumer demand improved, particularly in notebooks. Demand in China improved, and is looking better in the U.S. and Europe. Business demand is expected to rebound next year.
The results include a net favorable impact of $54 million ($0.08 per share) from a $66 million gain from a debt repurchase.
The GAAP operating loss was $77 million. AMD Product Company had non-GAAP net income of $2 million and operating income of $47 million.
$214 million adjusted EBITDA.
A number of new products were launched in the quarter, including the ATI Radeon HD 5000 family of graphics processors, Eyefinity multi-display technology, and server chipsets. New notebook chips already have 70 models planned based on them.
Revenues by segment: Computing Solutions $1.07 billion; graphics $306 million; foundry $256 million, other $21 million. Intersegment transfers eliminates $256 million of revenue. Unit volumes were higher across the portfolio. Computing and graphics segments both had double-digit sequential revenue growth and both were profitable. Chipset business had a record quarter; notebook, desktop and server all grew.
Ground was broken on the GLOBALFOUNDRIES fab in New York State.
Cash and equivalents balance ended at $2.51 billion, offset by $5.28 billion in long term debt.
Cost of sales was $811 million, leaving gross margin of $585 million. Operating expenses included $420 million for R&D, $221 million for marketing, general, and administrative, $17 million for amortization of intangibles, and $4 million for restructuring. Leaving an operating loss of $77 million. Interest net loss $110 million. Other income $47 million. Income tax benefit of $5 million.
Seasonality in guidance? Normal Q3 to Q4 is 9%, but there was a big build in Q3, so we are guiding to a moderate increase, less than average, 5% to 7%.
Gross margin contributions going to Q4? Q3 improvement from 45 nm, utilization, and better ASPs. More improvement from 45 nm in Q4, should have higher utilization. ASPs are a wildcard, but we think our products may be well-received. We still have a lot of capacity to take advantage of increased unit demand.
Separation of GLOBALFOUNDRIES and AMD Product Company? The biggest milestone will be third-party revenue for the foundry. It has to be a meaningful number, above 10% of revenue.
ASP trends? For CPUs, the biggest driver was desktop mix. In notebooks ASPs were flat. Server ASPs were up a bit. GPU ASPs were down slightly because the 5000 series did not contribute meaningful revenues in Q3.
Netbook offering? Our customers have netbook-like offerings already. Next month we will tell you more about our small form factor plans.
We are not hearing about any pileup of inventory. But to know we are in a normal cycle, we have to see the sales in Q4 of products built in Q3.
Faster growing, graphics or CPUs? Microprocessors usually grow faster Q3 to Q4. Usually Q3 is the best graphics quarter.
Europe? It is important for us, we have had good market share there, with a high attach rate for discrete GPUs. As Europe improves that should help us.
Congo and Istanbul? Our customers want two strong suppliers. Istanbul Opteron is doing well for us in 4 and 8 socket servers and in datacenters. Congo has a lot of design wins for thin and light notebooks. 6 core chips were about a third of our server mix.
Source of notebook strength? We did start selling Tigris components to OEMs towards the end of the quarter. We saw a very strong back to school cycle with the Puma chips.
We want to get 35% to 45% margins for graphic chips, 5000 series should help us get closer to that. Workstation graphics is a big opportunity for us, but it will be a slower introduction.
Congo budget thin and light notebooks for Q4 will be between 10% and 50% of mix. Congo notebooks release after October 22.
Corporate replacement cycle plans? We have 3 access points for enterprise market. Opteron. Commercial SMB type platforms. Workstation graphics. We are well-positioned for a commercial sector rebound.
Capital spend at GLOBALFOUNDRIES is mainly for AMD, but going forward will be increasingly for third-party customers.
32 nm sampling in first half of 2010, shipping to customers in second half.
Graphics demand, competition? We feel pretty good about 5000 line. You can buy ours today, performance and features are killer. OEM design ins are strong. Discrete GPU attach rates may be increasing.
Believes OEMs have drained supply chain in front of Windows 7 release. Just does not know what consumer appetite there will be for new computers.
HD5000 family is fully on 40 nm. Supply is short so far, but we expect it to improve in the coming months.
More cores, road plan? We will update you at the analyst conference next month.
Fusion technologies will sample in first half of 2010, shipping in second half.
Process generation gap? We started shipping 45 nm in Q4 last year. Economy delayed our transition because we had to sell out 65 nm inventory. 32 nm ramp should take place in two quarters, a typical ramp.
In our supply chain capacity utilization has increased dramatically in the last 90 days, but we have no direct examples of double ordering.
More Analyst Conference Pages:
Disclaimer: Our analyst summaries may include both our condensations of statements made by company representatives and our own analysis. They are not covered by any warranty. We cannot guarantee anything said by company representatives is true. We try not to make errors, but it is possible. Before making or terminating an investment you should always verify any factual basis of your decision.
Copyright 2009 William P. Meyers