Low Speed, Continued
In its response to rumors of further delays, AMD pretty much confirmed that, but said that it was going to put out only low-speed tri-cores and even lower-speed Phenoms because OEMs wanted it that way.
Moving on, the initial B3 chips are going to be low-speed Barcelonas.
What should you think about this? You should think that this is what a company does when it can’t make many medium- to high-speed chips.
I’ll put it another way, these guys can’t get a 2.4GHz K10 out the door, and aren’t even promising one for another three-five months. In comparison, Intel has had a 2.4GHz, 65nm chip out the door for a year (and it will outperform the 2.4 Phenom whenever that comes out).
Intel’s done a few other things since then, so for at least the first half of the year, it can rest assured that its lowest-end Penryns will outperform AMD’s highest-end K10s on the desktop.
That won’t bother the OEMs too much with their budget boxes, provided the price is right, and that’s going to have to be pretty low, if not quite as low as current X2 prices.
The bottom line is these chips will leave AMD a bit less financially sick, but they certainly won’t get it back on its feet.
Hope springs eternal, so the attention will shift to 45nm. For now, well, the chip doesn’t even exist yet. First silicon won’t be out for a few more weeks.
The plan has AMD getting this chip from 45nm to starting production in five months. Can they do that? Only if everything goes right, but that brings up another issue: What does “right” mean these days.
Based on some of AMD’s choices the last few months, I suspect “right” has become “Getting something out the door that works on time,” rather than “Getting a competitive part out.”
I suspect AMD needs 45nm more to offset the production capacity drop due to the mothballing of Fab30 and to lower the production price than anything else, and they’re leaving performance in God’s hands.
After getting down to around $5.50, AMD stock has jumped up almost 15% the last two days because the Attorney General for the state of New York has decided to begin an antitrust investigation against Intel.
How important is this? Not very, this seems to be a “Me, too” investigation for the following reasons:
isn’t doing too well politically as of late, and the current AG would certainly like to be governor some day, he’s run for the job before
anything serious, the EC will have come up with a ruling, and the AMD lawsuit will be close to trial if not actually being tried. Any New York State effort will mostly piggyback on these cases.
As we’ve said before, based on differing antitrust laws, it’s more likely that the European Union will look favorably upon AMD’s claims than the U.S.. The New York investigation is another iron in the fire, but it doesn’t change the fundamentals of the situation.
The New York investigation looks to be more of an opportunistic placeholder for an easy win down the road should AMD do well elsewhere than anything else.