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: Joe NYC who wrote (39197)	10/13/1998 3:20:00 PM
From: Steve Porter	Read Replies (1) \| Respond to of 1573718

Joe, Well for memory intensive apps, with a data size of say 24K (i.e. data that's being worked on), the K7 will be a big loser if it has a 2 cycle penality, when compared to say a PII, MII or even K6. Your data set has to be bigger than the 32K currently afforded by the currently L1 on the K6 to not suffer a 50% penalty. For example Mutant Alien Blaster applications may actually suffer, because the transformations that are usually done in 3D calculations usually operate on a small data set than would normally reside in L1 cache. Assuming the low latency of 3DNow instructions (say most taking 1 clock to complete), you are looking at a K7 performance being the same as the K6, and the only reason it's the same is it has an extra FPU to help with the work (i.e. 50% more processing to offset 50% latency penalty). Steve