Xref: utzoo comp.lang.c:11344 comp.arch:5515 Path: utzoo!attcan!uunet!lll-winken!lll-lcc!ames!amdahl!nsc!voder!apple!baum From: baum@Apple.COM (Allen J. Baum) Newsgroups: comp.lang.c,comp.arch Subject: Re: Self-modifying code Message-ID: <14017@apple.Apple.COM> Date: 15 Jul 88 16:46:51 GMT References: <3353@cognos.UUCP> <619@goofy.megatest.UUCP> <429@uwovax.uwo.ca> <33518@yale-celray.yale.UUCP> <1087@ficc.UUCP> Reply-To: baum@apple.UUCP (Allen Baum) Organization: Apple Computer, Inc. Lines: 14 [] >In article <1087@ficc.UUCP> peter@ficc.UUCP (Peter da Silva) writes: >I have a question: > > Why are an Icache plus a Dcache better than just > a big shared cache as big as both? In terms of hit/miss ratios, a unified cache is clearly better. However, in terms of effective access time, a split I/D cache is better. This is because both can be accessed simultaneously, instead of one having to wait for the other to finish. If both are getting accessed simultaneously (which should happen 40% of the time, if Loads/Stores account for %40 of instructions), then this more than offsets the increase in miss ratio. -- {decwrl,hplabs,ihnp4}!nsc!apple!baum (408)973-3385