AMD's Ruiz: First Quarter a 'Blip,' But Growth Estimates Foggy

More about Ruiz’ comments follow our original story from this morning:

11:52 am March 5, 2007 - With only about 50% of the revenue data for the fiscal first quarter of 2007 in hand, AMD had enough information ahead of a critical analysts’ meaning this afternoon to warn that it won’t be meeting its revenue targets this quarter.

Late last January, the company predicted its first full quarter with ATI as a component would be only seasonably lower than the $1.77 billion it collected during the final quarter of 2006. But even then, the company’s president and COO, Dirk Meyer, proclaimed, “We are not satisfied with our financial results in Q4.” That, coupled with a partly extemporaneous, heartfelt speech from CEO Hector Ruiz, gave a clear preliminary signal to the investment community that all was not well.

AMD has already taken a hit in server CPU sales in the last quarter, precipitated by a resurgent Intel. But its unit sales in other segments, including desktop and performance notebook CPUs, were actually quite respectable. It’s margins that are hurting the company, especially since last July, after Intel’s introduction of Core 2 Duo processors at very competitive prices triggered AMD to make price cuts in the midrange of its product lines.

Last January, AMD offered analysts guidance that revenue for this current quarter would come in at a range between $1.6 - 1.7 billion, with expenses estimated at $710 – 760 million, and about $120 million in projected charges.

With repeated references to his company’s “one-laptop-per-child” program as indication of everything that the economy is recovering, to that unit shipments will resume their growth rate, to that the sun’ll come out tomorrow, AMD CEO Hector Ruiz broke with his usual tradition of straightforward and candid comments. Instead, he filled the first five minutes of his interview before a Morgan Stanley analysts’ conference in San Francisco with vague and random references to his company’s history, past performance, and commitment to value.

It took Morgan’s lead analysts, and questions from the audience, to help bring Ruiz back to his usual form, but it took time and effort. In response to a flat question regarding something relatively simple like his predictions for the growth rate of the PC industry as a whole over the next five years, Ruiz began a kind of dance that he doesn’t perform all that well. “We are bullish on the way we see the growth opportunities for us,” Ruiz’ initial response began, “and we don’t measure ourselves against our competitor as we do against what we think we can do ourselves.”

Perhaps speaking over the amazement of the audience, wondering in what universe they had just awakened, Ruiz continued: “We look at the expansion of computing into the home, for example. There is no question that there is a gigantic potential of what a computer can bring into the home, not only in the area of entertainment, but in the area of communications in the home, in the area of managing technology, managing energy, bringing about a lot of the things that would make a computing capability invisible – where you get to the point where you know it’s there...” From there, he began talking about how homeowners tend to ignore such statistics as how many light bulbs they have, and how many are turned on.

After three minutes passed, the question was restated and resubmitted. Ruiz responded by saying growth should make the “high end of expectations,” as high as 20% per year – just not this year. Does this mean Ruiz really believes the PC market – whose present growth rate is believed to hover under 10% annually – will double in four years’ time? “I don’t see why it couldn’t,” Ruiz responded.

Finally, it came time to face the bad news head-on: Why was the company giving a warning that this quarter’s revenue would fall short of guidance? “[It’s] good news/ bad news,” Ruiz began. “In a very short period of time, we’ve gone from being...a significant player whose vast majority of products went to the channel distribution, not the OEM channel...to the point where now, the majority of our products go to OEMs and less to distribution.” Dell was perhaps a major catalyst of the ratio flip.

“That sort of transition occurred, in our view, probably faster than we had planned,” he said. In 2006, OEMs were expecting to be able to acquire more AMD products; and in responding to them, AMD was “not able to serve the distribution channel. As a result of that, when a lot of the OEMs were not able to materialize their demand as the year was ending...we couldn’t flip our capabilities to the channel as rapidly as we would have liked.”

At the same time, new OEM customers may not have been able to sell AMD-based products to the end customer as rapidly as they had anticipated. So AMD could not recover its expenses for having implemented the “flip” as rapidly as it had hoped.

Toward the end of the session, Ruiz was asked to give some clarity to AMD’s transitional roadmap from the current 90 nm generation through to 65 nm (where Intel is now) clear through to 45 nm. Ruiz would not go into any specifics on 45 nm, though he did shed light on 65 nm. “We have one factory that is, in terms of starting capability, on the material that gets started in the factory...our 300 mm factory is now at 65 nm. We have Fab 30 [in Dresden, Germany] which we’re in the process of transitioning on the diameter of the wafer upwards [from 200 mm].” During that transition process, he said, Fab 30 will be run outside of what AMD calls “the technology node” – meaning it may be doing testing and research work, but not production until the processes are refined there.

“We are in the process of moving the foundry to 65 nm over this quarter," Ruiz stated - by now, much more clearly, "and so we are pretty aggressively down this path of technology.” All of AMD’s wafer starts will be at 300 mm, 65 nm, he said, “this summer.”

How much better performance should we expect to see from this next "Rev. G" generation of processors? Given AMD's current on-the-record opinion of benchmarks, perhaps we shouldn't have expected the answer to that question to have come too easily for Ruiz. Still, Ruiz expects a 30-40% improvement in terms of benchmark scores for the server-based versions of AMD's "real quad-core" and other 65 nm dual-core products. But the success of that new product line, he implied, may be measured by whether AMD manages to finally "break the monopoly" of his competitor - whose name he rarely utters. Twice during today's conference, Ruiz said AMD needs to move from today's estimated 25% market share to 30% or better across the board, in order to achieve that monopoly break.