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: wanna_bmw who wrote (162195)	3/14/2002 2:26:04 PM
From: Dan3	Read Replies (1) \| Respond to of 186894

Re: I don't buy the bullsh!t about it fitting in the cache. The IBM 1.3GHz Power4, widely recognized as currently the most powerful CPU in both integer and floating point, has 128MB of L3 cache You've been awfully cranky lately - didn't start buying on margin when Intel hit $34, I hope. L3's are often quite a bit slower than L1/L2, and almost always off-die. Sometimes they aren't a heck of a lot faster than main memory. There is no way that the USIII hardware could have suddenly caught up and surpassed Power4 on this benchmark by such a wide margin, simply by adding compiler enhancements Actually, it's quite possible. SPEC contains some pretty bogus benchmarks. Bogus, because SPEC requires the use of some known bad (in terms of performance) algorithms that are not allowed to be corrected in the source. SPEC does allow for a compiler to recognize the known bad algorithm and replace it with "optimized" code - but it must be done in the compiler, not by correcting the source. So the "new compiler" is actually outputting a different program, in terms of the workload placed on the CPU. Since the performance problems and solutions of the SPEC source are well known, it's easy for a compiler writer to know what to search for and replace. Companies like Intel, that can afford to produce a "SPEC only" compiler, love this arrangement.