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 : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Windsock who wrote (167517)	7/7/2002 6:15:02 PM
From: Dan3	Read Replies (2) \| Respond to of 186894

Itanium Speculation: The Itanium II appears to be a chip design with some problems. An obvious part of the problem is due to it being made on Intel's old .18 Aluminum process, which slows it down. The more conductive copper interconnects used in virtually all other CPUs in production today, provide better performance by delivering higher voltage and current into a smaller core - which permits shorter paths and higher switching speeds. Just about everything else Intel makes has been shifted over to the faster, copper process - even Celeron! Why would Intel stick with the old, current limiting, Aluminum process for Itanium, the second generation of their flagship chip? They could be running into the same issues AMD is running into with Thoroughbred vs. Palomino (AMD's most recent .13 and .18 cores). The .13 process uses somewhat less power, but the die is much smaller, and it's hard to remove heat from a small die. So far, Thoroughbred is only a little faster than Palomino in standard boxes, but when both chips are supported with high performance cooling, Thoroughbred performance increases much more than Palomino's, indicating that the chip's core is quite a bit faster, but performance is being limited by the rate at which heat can be removed from the very small die. Itanium produces a lot more heat than any other chip with high volume expectations, and it may be that Intel is having trouble getting that heat off the die, even at the current, large, .18 size with low current aluminum interconnects. Their tests could have shown them that a smaller .13 die would be thermally limited to a lower performance level than the current .18 chips. AMD learned of the thermal issues of production copper chips in the late 90's, since it implemented a copper production process years before Intel did. Since then, AMD has been trying various solutions, including low-K, high-K, isotopic silicon, Silicon-On-Insulator (SOI), etc., and they claim to have finally found a solution last year with SOI, which they're putting into production this year. Intel, which until last year, hadn't experienced the need for SOI that arises when copper interconnects are used in production chips, won't be able to put SOI into production until 2004-2005. So, Itanium may be in something of a process box at this point. Itanium needs to be shrunk to .13 on copper to improve performance, but generates so much heat on that process that it must be run nearly as slowly (or even slower than) chips FABed using older technology. There is some hope; maybe Intel can contract out SOI Itanium production to AMD's more advanced FABs.... <VVVBG>