Monday, June 9, 2008

AMD Revival?

Are AMD's troubles mainly behind it? If they are, the stock price is way to low. But after watching Intel (INTC) swat down AMD with aggressively priced but inferior products, and the subsequent AMD fumbling of the transition to Barcelona quad-core Opteron processors, I can understand why confidence in AMD is not high and could take some time to rebuild. I kept the AMD stock I accumulated, but I have not yet added more to my position despite recent tempting prices.

There are three big variables to look at. One is the competitiveness of AMD's technology. This has many moving parts, including many moving parts at rivals Intel and NVIDIA [NVDA]. The second is financial: what does the hard, historic data look like and tell us about future potential. The third is about markets. A number of nations have accused Intel of using illegal tactics to keep AMD out of the market. And many well-place decision makers own Intel stock. And Intel advertises heavily in the trade journals that help technology purchasers make decisions. As in other businesses, having a better product is no good unless you can develop channels for selling that product.

I think the best way to see the financials is to look at past quarterly reports and analyst conferences. Of course you can look at SEC filings, but for a comprehensible overview I rely on my own analyst conference summaries. Follow this link for a summary of AMD's April 17, 2008 report on the quarter ending March 1. AMD's investor relation page is also useful.

The biggest change over the last year is the decrease in the write-down of ATI acquisition charges. Graphic chip maker ATI was barely holding on against an NVIDIA onslaught when AMD purchased ATI for about $5 billion. Today the combined AMD/ATI operation has a market capitalization of about $4.5 billion. Hence the write-downs. AMD's Q1 cash flow from operations was $16 million, and they generated $50 million more by selling obsolete (to them) tools. More interesting, revenues were $1.505 billion, down 15% sequentially from $1.77 billion, but up 22% from $1.23 billion the year-earlier quarter. The seasonal drop was a bit worse than expected, but the 22% pop from year-earlier seems to show the worst of the revenue slump is behind.

I don't expect anything beyond the normal seasonality for the June quarter. Quad-core Opteron sales should be ramping up, and quad-core Phenom too, but neither have enough technological advantage over Intel's quad-core products to cut much into Intel's market share. The same is true on the discrete graphics processor side, where NVIDIA's domination seems solid in the short run.

Looking down the road 6 months to a year, however, I think AMD is going to be able to leverage its technological leadership over Intel in some key areas. AMD will continue to be the low-energy leader, which is especially important in servers and notebook computers. While Intel has increased the efficiency of its chips (largely by copying AMD) and has a 6-month to a year lead in process technology (the constant shrinking of transistor size on chips), its designs still seem to ignore the needs of end-customers.

The single biggest differentiator, however, is in graphics. AMD's Puma [Turion X2 Ultra] technology for notebooks is the best example. Intel cannot make high-end graphics chips and its low end chips are garbage (and the main reason people complain about Microsoft Vista). AMD Puma builds in a high degree of graphics functionality at a low cost, ideal for low-end notebooks. But for more expensive notebooks, manufacturers can simply add a discrete, high-end, graphics chip. The graphics processing power of the two chips is additive.

Everyone is the industry is afraid of Intel. If Intel cuts off your supply of chips, you are screwed. If Intel raises its prices to you (or gives your competitors better discounts for not using so many AMD chips), you are screwed. There are many large corporate end-users that still have Intel-only policies. Consumers usually do not know what to think about AMD v. Intel, although that is an improvement over 10 years ago, when AMD was assumed to be the inferior brand.

On the other hand, if someone else is selling notebooks that are AMD based (AMD makes the processor, graphics processor, and "glue" chips) that are noticeably better (better graphics, faster processing, longer battery life) than more expensive Intel-based notebooks, that could cause some people to rethink their relationships with the bully boys at Intel.

Of course, the Intel camp is promising wonderful, better than AMD notebook chips real soon. And they will buy enough advertising (and reviewers) to make their propaganda seem true, whether it is or not.

So AMD remains a risky stock. It has a lot of potential. But so far betting on the PC industry showing some backbone against Intel bullying has not paid off, and it may not in the future.

I think Intel is going to have to pony up some money to settle the AMD v. Intel lawsuit. There has just been too much evidence accumulating around the world of Intel wrongdoing. Whether it will be for a half-billion dollars or for $50 billion or any point in between is anyone's guess right now.

No comments:

Post a Comment