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: Ali Chen who wrote (119131)	11/24/2000 4:56:48 PM
From: Dan3	Read Replies (1) \| Respond to of 186894

Re: why we do not see any SPECfp benchmark results on dual P-III or quad Xeons? First let me state that I don't think adding a second processor is anywhere near as effective as having a first processor that is twice as fast. But if the clawhammer core is 20% better, clock for clock due to better branch prediction and instruction set optimization (SSE etc.), then I still think it's possible that the second core could take it up another 25% to a 50% increase. Having said that, let me offer a quick reminder that SPECint_2000 and SPECfp_2000 are designed to benchmark single processor systems while SPECint_rate2000 and SPECfp_rate2000 are designed to benchmark multiprocessor configurations. SPECint_2000 isn't used to benchmark multiprocessor systems because they wrote it for uniprocessor systems while the SPECint_rate2000 is the version for multiprocessor systems. So (as expected) there are almost as few single processor systems listed as tested for SPECint_rate2000 as there are multiprocessor systems listed as tested for SPECint_2000. spec.org spec.org But there may also be a real difference in performance of a system where two die on a core share an L2, compared to conventional dual processor systems where each core has its own L2. I don't have the link handy, but some work done profiling X86 application performance showed a big drop in dual processor systems working on code/data that fits in the L2 vs. single processor systems. Absent the need to bus snoop the L2, Sledgehammer may perform better than conventional dual processor systems. On you point regarding the very small benefits to be expected of 64 bit vs. 32 bit operation for the vast majority of applications. I agree that it will, at best, be an occasional help, and sometimes even a hindrance. From personal experience I can say that we have a lot of dual processor SUN and Intel systems at work, and they do seem to do somewhat better then similarly clocked uniprocessor machines. Regards, Dan