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 : CYRIX / NSM
NSM 18.270.0%Jul 31 5:00 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Craig Freeman who wrote (30319)11/20/1998 3:14:00 AM
From: Scumbria  Read Replies (1) of 33344
 
Craig,

1MB of L2 cache is about the all you see on Socket 7 mobos. Why?

When you go above 1MB, the performance improvement (in single processor systems) tails off very quickly. Caches are only effective for reuse of code and data. The first time a code or data block is loaded, it will miss in the cache. Similarly, context switches incur a large percentage of cache misses. Even with an infinite sized cache, a certain fixed percentage of memory accesses would miss in the cache.

As clock speeds increase to 1GHz and above, high performance systems will have to use more sophisticated solutions than traditional sdram or rdram. There are other types memories available, i.e. edram, sram, etc. which computer architects will have to take advantage of.

Scumbria
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext