SI
SI
discoversearch

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: Scumbria who wrote (80640)5/9/1999 11:09:00 AM
From: sridhar ramasami  Read Replies (1) | Respond to of 186894
 
To all, surprised at thread.

I am shocked that you guys are discussing relevant and technical matters instead of slinging mud at each other like a normal thread.

What is the world coming to ?

Sridhar



To: Scumbria who wrote (80640)5/9/1999 1:06:00 PM
From: Paul Engel  Read Replies (3) | Respond to of 186894
 
SCUMbria - Your cache performance calculations are even worse than your investment choices.

Intel's internal testing has shown that a 533 MHz Coppermine outperforms a 600 MHz Pentium /// with off-chip 512K 1/2 speed cache.

June is only 22 days away - how far away is the desKtop 7 ?

Paul




To: Scumbria who wrote (80640)5/9/1999 7:01:00 PM
From: Tenchusatsu  Read Replies (1) | Respond to of 186894
 
Scumbria, check out Intel's performance brief on the Pentium II Dixon:

developer.intel.com

As you know, Intel replaced the 512K of off-chip L2 cache in the Mobile Deschutes with 256K of on-die L2 cache in Dixon. This led to a 5% speedup in Business Winstone 99. SpecInt95 was even better, a 7% speedup, and SpecFP95 saw a 13-14% speedup. However, MultiMediaMark 99 saw no difference and 3DMarkCPU saw only a miniscule improvement.

So in essence, the benefits of moving from a 512K off-chip L2 cache to a 256K on-die version will vary widely with the application. Toss Rambus and AGP into the equation, and you'll have quite a complex performance model.

Tenchusatsu