Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: jackthetab who wrote (262546)	11/12/2009 9:32:01 PM
From: Elmer Phud	Respond to of 275872

Doesn't answer the question of what's the viable and profitable (low cost) alternative. I think you made your own point. There doesn't appear to be a better choice considering they're all running basically the same process anyway, with AMD, IBM and chartered adding an SOI derivative. Something you need to consider. All foundries are compromises. They can not tailor their process for a single customer. A processor made at a foundry will be at a disadvantage performance wise, all else being equal (which is never the case either). This will likely take AMD out of the highend market but perhaps lower power applications might be more competitive if they can come up with a core like Atom. Not only is Atom low power but it is extremely small allowing thousands of die per wafer. AMD can not currently compete in that space because of die size alone. Their costs are just too high and GF, or someone else, will have to get their profit as well.

To: jackthetab who wrote (262546)	11/12/2009 11:40:06 PM
From: fastpathguru	Read Replies (2) \| Respond to of 275872

Doesn't answer the question of what's the viable and profitable (low cost) alternative. Perhaps the benefit is flexibility. Like I posted about Bobcat earlier, from yesterday's slides: Very low power design – Sub one-watt capable 90% of today’s mainstream performance in less than half of the silicon area * Synthesizable / Easy to Reuse * Complete ISA support – SSE1-3 and virtualization The design is done. It can be ported easily to any foundry's current or future process. It can be combined by AMD w/3rd party IP. AMD doesn't have to dedicate capacity from a bleeding-edge fab to pump out as many as they want; they only need the bleeding-edge capacity for the boutique designs. Before, mass-market vs. boutique designs were essentially competing for limited capacity within AMD's expensive high-end fab... I.e. non-optimal capacity utilization. (Chartered added only token capacity, being limited to 20% and running tweaked designs that required SOI.) Now, mass-market and/or customized x86 chips can be pumped out ad infinitum, without competing for the same capacity that will produce the boutique designs. That's a decidedly non-trivial gain, from my perspective. fpg