Where do you start?
Tech Report has now quantified the impact of disabling logic in the TLB. Not too pretty at best, and disastrous for Firefox.
Based on AMD’s own statement, Barcelona sales have been restricted to very specific customers using them in specific work that won’t trigger the flaw. Tech Report has heard those specific customers are getting a “substantial discount” for their troubles (and wondered loudly whether folks like us should get the same, too).
Even before this, though, AMD stock has not been doing well at all lately.
To rework an old joke, Hector Ruiz goes back to a college reunion and boasts, “In 2007, I made AMD a $5 billion corporation.” Someone asks, “What was it before that?” and a classmate now with Intel says, “A $10 billion corporation.”
Here’s a chart (courtesy of Marketwatch) from just the last month showing AMD’s stock performance (blue bars) compared the Dow Jones index (brown line):
Most of the decline has occurred in the last three weeks, beginning when that UAE place bought 8% of freshly-printed AMD stock.
They’ve lost almost 200 million dollars on their 600 million investment already. A 30% loss in just three weeks! If only you could make that into a weight loss program!
Seriously, though, how much harder will that make any future efforts by AMD to get additional funding any time soon?
What is strange about recent events is that they uncover other previously unknown bad things.
Just give just example, all these Barcelona problems aren’t going to hurt Barcelona sales. Even before these problems, mainstream OEMs didn’t want the initial ones because they were too slow.
“[Dean McCarron, head of Mercury Research, a PC analysis company] said major computer makers have told him they plan to begin volume production of Quad Core Opteron-based systems when AMD begins volume shipments next year and bumps its speed. “The initial launch speed wasn’t what they wanted in terms of speed. They made it clear when the higher speeds were available they would be more aggressive in ramping,” he said.
That was personally quite amusing since AMD said not so long ago that these same customers were demanding slow processors.
The scary part about all this is that all this isn’t even the worst of the news. The items listed above are spilled milk, they’ll hurt AMD’s financial results this quarter a bit, hurt its reputation and credibility even more, but nothing deadly.
It’s the following news that will likely hurt even more:
First, though this is mostly symbolic, in the midst of all this news about slow processors getting even slower, maybe it wasn’t the best time to announce that you were phasing out your fastest processors.
Granted, they had to go since Fab30 is shutting down, but doesn’t that sort of point out the inadequacies of AMD’s 65nm process? Besides, when your company’s future rides on K10s, why are you coming out with eleven new K8 models over the next six months??? That doesn’t exactly sound like a vote of confidence in your ability to make K10s, does it?
Maybe that’s because AMD also announced that the B3 stepping, all-fixed 2.6GHz Phenom has a 140 watt TDP. Doesn’t look like B3 cuts down on the power, and it’s tough to see a 3.0GHz OEM chip coming from B3. Then again, AMD hasn’t promised more than 2.8.
But not even that strikes me as being the worst news. AMD can probably struggle through most of 2008 if they do what they currently say they can do, and the global economy doesn’t go seriously south in the meantime.
However, 45nm then becomes make-or-break for the company. 45nm has to come in on time, and it has to be good.
We’re supposed to hear more about AMD’s 45nm plans at the analyst conference December 13, but there’s growing signs 45nm won’t give AMD the boost it needsto get competitive again.
And that could be the final mistake.
I don’t know about you, but I asked Santa to give me some serious (and real) good news about AMD for Christmas. How about you?
Well, if it’s going to happen, Christmas will come on December 13, which will be the AMD analyst conference. Tomorrow, we’ll talk about that.