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: Road Walker who wrote (61991)	11/3/2001 10:11:41 AM
From: combjelly	Read Replies (1) \| Respond to of 275872

"I never said that." Maybe not in those words. But you have consistently and repeatedly held up clock rate as the "industry standard". "But to create a unilateral benchmark that mimics MHz numbers has nothing to do with reality." Well, based on the experience with SPEC, it is a natural thing to do. First, the SPEC ratings followed MIPS ratings. Then it chose to rate in multiples of a particular model of a Sun workstation. I don't know what they are using now as their base measure. AMD claims that their benchmark is against their Tbird. Assuming that is true, then their numbering scheme is fairly natural. They could have normalized it to something like 1.5+, 15+, 150+ or the chosen 1500+, only the center two avoid what you are complaining about. Sorry, I didn't catch your earlier post, you were taking a more rational stance then. If you feel so strongly about it, then why don't you write Intel and suggest that they do something like that? It would be a good counter to what AMD's quantispeed. Intel will need to do something fairly soon, else they will be having a lot of trouble justifying the low clock rate of Itanium... " I'm either "ignorant, stupid or deceptive"" Sorry, you are if you argue that AMD is "perverting the industry standard", i.e. clock rate. Either you never had the information about why clock rate is a bad metric, you got the information but didn't understand it, or you understand it, but choose to ignore it because it doesn't give Intel an advantage. There just are not a lot of other ways to view it. "With that pleasant thought in mind, I'll end this side of the argument, confident that you will be banned from the thread." There is a procedure, use it. But then you wouldn't have anything to whine about, would you?

To: Road Walker who wrote (61991)	11/3/2001 10:40:47 AM
From: Dan3	Respond to of 275872

Re: I'm either "ignorant, stupid or deceptive" What makes you think the set is disjoint? Just kidding! You have seen what they call me on the Intel thread... :-)

To: Road Walker who wrote (61991)	11/3/2001 10:50:21 AM
From: pgerassi	Read Replies (2) \| Respond to of 275872

Dear John: Since AMD is pushing TPI with Quantispeed as a transition to it, your complaints have no merit. Why aren't you blasting Intel with their iCOMP which does the same thing? It changes over time and is thus, even more arbitrary than the basis for Quantispeed. You just don't like that P4 loses a deceptive advantage. What the marketplace needs is a benchmark whose creation is truly independent of OEMs, is free to the public so that any claim maybe verified easily and any public advertising is controlled to verified results by third parties. The first keeps things like what happens with Sysmark from happening where the applications chosen are biased towards one OEM's newer products. The second makes things like JC's databases of system benchmarks able to be viewed by the public and done for no cost to the user to his machine unlike SPEC and many other benchmarks. The third is a possible way to pay for independence of an organization that monitors for "hanky panky" and verifies that the benchmark results are valid and repeatable. It also must pay for a peer review of the benchmarks created and thus show that the selection of a particular program or method is justifiable to the goal of showing performance of real machines for real application workloads in real environments commonly found in typical target customers. This keeps out the use of a little used program that very few people use over one that is commonly used and the way they use it (WME selection and usage in CC for example). This is how true standards like the measure of HP are done. There are very few benchmarks that come close to it. Pete

To: Road Walker who wrote (61991)	11/3/2001 2:50:21 PM
From: hmaly	Respond to of 275872

John Re..): Message 16454501 Of course if you rate the complete system performance, that takes the brand equity away from the microprocessor manufacturer, and transfers it to the OEM. Probably a bad thing for Intel, maybe a good thing for AMD.<<<<<<<<< I agree with your post of rating the entire system. Who cares how the individual company will benefit or lose, as long as the customer wins, and the customer will win as there are many ways to improve system performance besides buying a faster CPU. Boy, for a moderated thread, you AMD guys sure throw around insults<<<<<<< LOL Thats a good one. I accidently read about 20 posts on the Intc thread the other night, and my ears were blistered. That thread makes use look like a bunch of nice guys. And that was just Paul's posts.