Path: utzoo!attcan!uunet!lll-winken!lll-tis!ames!amdahl!pyramid!prls!mips!mash
From: mash@mips.COM (John Mashey)
Newsgroups: comp.arch
Subject: Re: Mips figures.
Message-ID: <2870@winchester.mips.COM>
Date: 22 Aug 88 19:23:12 GMT
References: <422@ma.diab.se>
Reply-To: mash@winchester.UUCP (John Mashey)
Organization: MIPS Computer Systems, Sunnyvale, CA
Lines: 25

In article <422@ma.diab.se> pf@ma.UUCP (Per Fogelstr|m) writes:
>Just wondering....
>
>Anyone else who has seen the connection  Dhrystones / 2 = Claimed mips ?.
>
>It work out for nearly all modern processors.

>Well, what the ...  I just thought it was a fun thing.

It's not a terrible rule of thumb, for scalar processors.
However, there is enough variation that you should be careful.
For example:
VAXen: around 1.8
MIP R2000 & HP Precision: both in 2.1-2.4 range
(I pick these because we have enough cross-correlations to believe
in consistent vax-mip (integer) ratings, and all have good compilers.)
Note that there is a 30% variation already from top to bottom.

In general, Dhrystone is a good benchmark to use to beat up on
VAXEN for relative-performance numbers, especially with a modern RISC
with good compilers.
-- 
-john mashey	DISCLAIMER: <generic disclaimer, I speak for me only, etc>
UUCP: 	{ames,decwrl,prls,pyramid}!mips!mash  OR  mash@mips.com
DDD:  	408-991-0253 or 408-720-1700, x253
USPS: 	MIPS Computer Systems, 930 E. Arques, Sunnyvale, CA 94086