Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ames!sun-barr!newstop!texsun!convex!eugene!swarren
From: swarren@eugene.uucp (Steve Warren)
Newsgroups: comp.arch
Subject: Re: How Caches Work
Message-ID: <1710@convex.UUCP>
Date: 11 Sep 89 14:16:17 GMT
References: <21936@cup.portal.com> <1082@cernvax.UUCP> <16306@watdragon.waterloo.edu> <RANG.89Sep10184900@derby.cs.wisc.edu>
Sender: news@convex.UUCP
Reply-To: swarren@eugene.UUCP (Steve Warren)
Organization: Convex Computer Corporation, Richardson, Tx.
Lines: 14

In article <RANG.89Sep10184900@derby.cs.wisc.edu> rang@cs.wisc.edu (Anton Rang) writes:
>  The problem is that caching something is not without cost.  Loading
>128 words into the cache from main memory is generally just as
>expensive as fetching 128 words from main memory directly.  The cache
>only helps if you're doing multiple fetches of the same memory
>location before it's overwritten with other data.

Not if main memory is pipelined and interleaved.  In that case the
cache may be filled fairly rapidly since consecutive locations are
accessed.

--Steve
-------------------------------------------------------------------------
	  {uunet,sun}!convex!swarren; swarren@convex.COM