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: Elmer who wrote (92725)11/17/1999 3:25:00 PM
From: Tenchusatsu  Respond to of 186894
 
Elmer, <Are you suggesting that the huge boost in FP
performance was more a result of the RDRAM rather than the
SSE?>

In the performance briefs, both the 440BX and the 840
systems were using Intel's V4.5 compiler plug-ins for both
C/C++ and Fortran. And both systems were using Coppermine
processors (since there are no Katmai processors that can
run at speeds over 600 MHz).

That performance advantage is definitely due to RDRAM, since
SSE is being used in both systems. By the way, here's the
chart again, but this time with an 820 system added:

440BX 440BX 820 820 840 840
650 MHz 700 MHz 667 MHz 733 MHz 667 MHz 733 MHz
------- ------- ------- ------- ------- -------
SPECint95 31.6 33.8 32.8 35.7 33.0 35.6
SPECint95_base 31.3 33.5 32.4 35.2 32.4 35.3

SPECfp95 22.9 23.5 26.9 28.1 29.0 30.4
SPECfp95_base 22.4 23.2 26.3 27.7 28.2 29.7

Once again, not much of a difference in SPECint, but a
sizeable difference in SPECfp. Seems like SPECfp is very
dependent on memory bandwidth.

Tenchusatsu



To: Elmer who wrote (92725)11/17/1999 3:26:00 PM
From: NITT  Respond to of 186894
 
RE "Are you suggesting that the huge boost in FP performance was more a result of the RDRAM rather than the SSE"

I'd suggest it's the big internal cache along with some of the better plumbing in the CuMine... any real engineers out there that can help me with this idea?

Nitt