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.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: Paul Engel who wrote (43729)	12/21/1998 2:00:00 PM
From: Tenchusatsu	Read Replies (1) \| Respond to of 1573029

The importance of L2 cache in Winstone 99 ... Just looked again at my PC Mag Dec. 15 issue, and it had a comparison of home PCs. In the Business Winstone 99 chart, here are the scores of two Gateway models: GX450-XL (Pentium II Xeon 450 MHz): 25.2 G6-450 (Pentium II 450 MHz): 23.5 As far as I can tell, both the G6-450 and the GX450-XL are similarly configured, with 128 MB of SDRAM, identical hard drives, and Riva TNT cards. The chipsets are different, 440BX for the Pentium II model, 440GX for the Xeon, but in my mind those two chipsets are identical. The Xeon model had more features, of course (e.g. hardware MPEG-2 decoder, better audio), but none of them contribute to a higher Winstone 99 score. Now as we remember, the Xeon showed no advantage over the Pentium II when it came to Winstone 98. Now, we see a 7% increase in Winstone 99 thanks to that full-speed L2 CSRAM cache. So what does this mean? Two things. One, we know that the Celeron with 128K of on-die L2 cache matches the performance of a Pentium II with 512K of half-speed L2 cache. This means that Dixon and Coppermine, with 256K of on-die L2 cache, will do even better in Winstone 99. (And let's not forget that Cascades, with 1 MB of on-die L2 cache, won't look too shabby on the benchmark as well, but of course, that CPU is for servers and high-end workstations.) Two, could it be possible that the K7, with its L2 cache off-chip and running at 1/2 to 1/3 speed, scores rather low in the Winstone 99 benchmark? Remember that Anand's K6-3 setup not only had 256K of on-die L2 cache, but also a whopping 2 MB of motherboard L3 cache. I don't think the K7 will have any L3 cache on the motherboard at its release, but of course, things change quickly in the PC market. Tenchusatsu