AMD Press Conference. . .

Most Memorable Jerry Sanders Quote of the Conference: “We think the Duron can compete effectively against the Pentium III. We think the Pentium III is dead meat and the only weapon they’ve got left is price.”

What You Need To Know

Financial

Financial prospects the way Wall Street measures them probably not all that hot, because the taxman is back! AMD expects to have pretty decent growth (pretax earnings going from about a $2.50/share pace to a $3.00/share pace for 2001. However, most of that growth is going
to end up with the IRS.

This is something a lot of people forget about recovering companies; in the early stages of recovery, they pretty much get a free ride
from Uncle Sam by offsetting their initial profits with previous losses.

AMD’s effective tax rate will go from 15% last quarter to an expected 32% for 2001. They expect to earn about $2 a share posttax for 2001, which is roughly what they’ve been getting lately.

Another fab is definitely in the works for 2004. AMD definitely needs another one to turn this from Hertz/Avis to Coke/Pepsi.

A minor but very indicative note about the AMD figures; they’re still running a tight ship. They kept a tight rein on non-manufacturing expenses, which often bloat when companies are having good times. Far more impressive than, say, Apple.

Technical

AMD will come out with a non-Palomino 1.3Ghz chip in next couple months. (I wouldn’t take 1.3Ghz literally; it could well be 1.33Ghz, and AMD is just rounding off.)

By the end of the first quarter, all TBird will be manufactured in Dresden at 1Ghz or better.

Aluminum Durons will continue to be made in Austin, AMD expects to be able to hit 1Ghz by 3Q with them.

Palomino comes out during the 2Q 2001. Supposed to have a few improvements that help performance “a bit.” Starts off at 1.4/1.5Ghz; 1.7Ghz second half of the year.

They don’t expect to reach 2Ghz until Clawhammer in 2002. .13 micron migration at Dresden pushed up a little to 4Q ’01.

They don’t expect a rapid turnover to DDR either: marketshare of 20-30% by the end of the year.

AMD has stopped making K6 family wafers. K6-2 sales dropped to 2 million last quarter, will be negligilble this quarter.

Observations:

These were not the forecasts of a company that expects to aggressively grow in 2001, but rather the comments on a company intent on consolidating its gains and maybe grabbing a little more during the year.

They don’t expect a price war, which doesn’t mean there won’t be one, but is a pretty good clue AMD won’t start one. Why should they, they already took their shots. Their anticipated ASPs for 2001 ($90-$100 in the post K6 era) indicate to me that we’ll see pretty much the same price structure for their CPUs in 2001 that we have now, which isn’t bad at all.

From the comments on Palomino rampup, seems to me the overclocking strategy will be to upgrade earlier rather than later, and shoot for about 1.7Ghz or a little better.

Sure didn’t seem like Durons/Morgans were going to get a whole lot better during 2001, maybe a bit over 1Ghz, but some of you are there already.:)

Competition from Intel? Willamette? Please, even the mainstream media can’t even pretend it’s good. Tualatin? That chip has never made sense to me. Seems like too little, too late, and probably costing too much. If Intel still plans on a Coppermine-X that will work on current mobos, that may be a good choice for some.

We’re supposed to see an SDR board for Willy, but I think that’s going to go the way of the 810 board, a crippled platform. If Willy doesn’t do too well even with a dual-channel RDRAM crutch, how bad is it going to be with PC133?

No, Palomino still looks like the only game in town for a major 2001 upgrade. Second half of the year, we’ll probably see incremental improvments as DDR memory gets better. There’s only two real potential factors I see that could upset the apple cart.

The little one towards the end of the year might be a dual-channel equivalent DDR (might be really dual-channel, might be some quad-pumped thingie).

The big one is Intel panicking, at any point in the year. As I predicted a while back, AMD has positioned itself so that Intel is essentially the shock absorber for world demand, at least for the next couple quarters.

Intel is expecting CPU demand to be down 15% next quarter, which is more than the usual seasonal downturn. If the economy heads south, it could get a good deal worse.

More ominously, at least from the stock market side, Intel isn’t going to be able to prop up its earnings with stock sales. They’ve sold the jewels already. Intel’s financial statement states that their financial arm as the end of 2000 only has $290 million in unrealized gains (i.e. what they would make if they sold the stock). That’s less than the $450 million in gains they reported last quarter. Unless the tech sector makes an amazing comeback, that source of profits is gone.

This is what I suspect the picture is Inside Intel:

“We’re stuck for the next year. AMD’s got us where they want us the next couple quarters. Thank God they don’t have fabs. We got talked into ripping too much out of Willy (Son of Cacheless Celeron?), and now we have to redo the chip to get them back in again. It’s so much of a pain that we have to resurrect the PIII to save our asses for the second half of the year. Marketing people, earn your pay! If the Blue Guys can’t do it, go find some Green and Purple Ones, too.

And just whose idea was Rambus, anyway?”

The proper strategy for Intel is to grit its teeth, give up some market share, maintain price structure, and ride out the storm for 2001. Will they be disciplined enough to do it, or will they panic? For now, I think they are. Six months from now; I don’t know. If they do panic, all bets are off.

My advice to you if you want to upgrade in 2001 is to look to Palomino, probably early summer for many of you. Go to DDR either sooner or later depending on your individual situation. After that, then open up a bank account and put your computing mad money in there, because your next good shot won’t happen until 2002, probably not until well into 2002. Might be Clawhammer, might be .13 micron Willy, way too early to tell.

Email Ed



Be the first to comment

Leave a Reply