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: peter_luc who wrote (72939)	3/2/2002 7:33:30 AM
From: Gopher Broke	Read Replies (1) \| Respond to of 275872

However, Intel will finalize its decision about including or not including Yamhill technology into the upcoming Prescott core a bit later, depending on the success of the coming AMD Hammer competitors. What a joke. There is no pressure, Intel can simply decide about x86-64 later, when they have seen Hammer. If Intel really are just planning to bolt an interpreter onto the front of Prescott then its x86-64 mode performance going to suck. AMD will be able to rate Clawhammer's 64 bit mode performance as 10,000+! It will be interesting to see Intel try and market their way out of this hole. They can probably hold their own while x86-64 is constrained to the Linux server market. (That is assuming they have a plentiful supply of P4s running over 3 GHz by the middle of next year, to counter ClawHammers 32 bit performance.) However, if Microsoft produce Windows/x86-64 before Intel has a full 64 bit core then they are in big trouble. Can Intel really persuade Microsoft to not produce Windows/x86-64 for the three years that it will take them to produce a true x86-64 core? I don't think so, given the ever-present threat from Linux. AMD are right when they say Microsoft can't affort to drag their heels on x86-64 support.

To: peter_luc who wrote (72939)	3/2/2002 9:25:21 AM
From: Dan3	Respond to of 275872

Re: even more evidence that Prescott will indeed use Yamhill technology: There are a couple of reasons why Intel will want to drag its heels on this one. The obvious one is Itanium - having misled the industry and been caught on Rambus, Intel will try to avoid doing the same on Itanium quite so soon. The next issue is that 64-bits on the desktop isn't that important for most applications. In fact, 32-bits on the desktop wasn't necessary for quite a few applications. What that means is that "tweaked for Intel's X-86" applications often won't run any faster on Yamhill than on on Hammer. Even more so than the 32-bit transition, the first few years of the transition to 64-bit hardware on the desktop is much more about marketing and being "ready for the future" than it is about near term performance gains. Intel may also see more, or less, of a die size increase from the move the 64-bits than AMD did. Exactly why the P4 die is so large is still something of a mystery. If Intel's chip suffers less bloat, then AMD will still probably have the smaller die, but if it's more, then Intel may be facing more wire delay and fabrication issues from its already "too big" die.

To: peter_luc who wrote (72939)	3/2/2002 9:48:58 AM
From: niceguy767	Read Replies (1) \| Respond to of 275872

peter: "Uh, what happened? I just woke up from a nightmare." Sounds like another "nightmare" screenplay imagined by Elmer!

To: peter_luc who wrote (72939)	3/3/2002 1:03:34 AM
From: Petz	Read Replies (3) \| Respond to of 275872

peter, re:<Yamhill in Prescott, disabled (for now)> even more evidence that Prescott will indeed use Yamhill technology: See this quote from x-bit labs: "We also have some info (proven by Intel as well) that Prescott may also support Yamhill technology, which will allow these CPUs to process 64bit codes. However, Intel will finalize its decision about including or not including Yamhill technology into the upcoming Prescott core a bit later, depending on the success of the coming AMD Hammer competitors." I am very dubious about Yamhill being included in the first version of Prescott to be released probably mid 2003. The idea presented above is that, if AMD's Hammer series appears to be successful, Intel will enable the Yamhill technology in it's first Prescott chips. To do this, Intel would have to abandon Socket 478 and develop a whole new set of chipsets capable of handling the wider address bus. Motherboards would also have to support address bits that may never be enabled. There is no chance that all of this could even be ready by mid-03. Another thing that doesn't make any sense is that, it will be at least March, 2003 before Intel even knows if Clawhammer is a "success." That is absolutely too late to get any kind of software support for a NEW instuction set, call it i86-64, before March, 2004. Therefore, the idea of deciding inclusion/exclusion of Yamhill technology based on AMD's success necessarily EXCLUDES the idea of doing anything except copying AMD's x86-64 exactly. The idea of expecting software support for a concept that may or may not be included is ludicrous, expecially if the decision won't be made for a year. Even getting motherboard makers to build products for a new socket and new chipsets, but not knowing whether to have a 32-bit address bus or a 40-bit address bus until a few months before product release is ridiculous. Petz