From: Kevin AT Quitt DOT net (Kevin D. Quitt)
Newsgroups: comp.os.msdos.djgpp
Subject: Re: Optimization
Date: Sun, 01 Dec 1996 17:05:07 GMT
Organization: Cruisin' from home
Lines: 17
Message-ID: <32b1ba0f.58325187@news.pacificnet.net>
References: <57hg9b$or5 AT kannews DOT ca DOT newbridge DOT com> <329C95AD DOT C3E AT silo DOT csci DOT unt DOT edu> <57k531$5bu AT kannews DOT ca DOT newbridge DOT com> <slrn59ubiq DOT nb DOT nxk3 AT b63526 DOT student DOT cwru DOT edu> <57nsm0$cvp AT lyra DOT csx DOT cam DOT ac DOT uk>
Reply-To: Kevin AT Quitt DOT net
NNTP-Posting-Host: 207.171.23.21
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp

On 29 Nov 1996 23:47:12 GMT, gpt20 AT thor DOT cam DOT ac DOT uk (G.P. Tootell) wrote:
>ok. i'm confused now. i thought the cache was 32 bytes but 128 bits is 16 bytes
>no? so just how big is the cache :) or did it change between the 486 and pentium?

The cache reads 4 words of 32 bits in each load.  The idea is to already have
what you're probably going to need next 75% of the time, rather than having it
there in case you need it again.  The P6 instruction cache, for example, is
sensitive to the instructions therein and will prefetch from both possible
targets of a conditional jump (before the instruction even gets to the
processor).  That's why absolute and conditional jumps are now considered to
take zero cycles on advanced processors like the P6 and 68060.

--
#include <standard.disclaimer>
 _
Kevin D Quitt  USA 91351-4454           96.37% of all statistics are made up
Per the FCA, this email address may not be added to any commercial mail list