From: broadley@neurocog.lrdc.pitt.edu (Bill Broadley) Subject: Re: MIPS R3000 board project; update. Date: 21 Jun 1993 06:09:44 GMT
On the mflops2.c benchmark (from marlin.nosc.mil pub/aburto) our 12.5
Mhz mips 2000 outperforms a 486-66 by a marginal amount. This seems
representative of the other fp codes I've run on the decstation.
We have a 16.67 Mhz mips 2000 (decstation 3100) which is proportionally faster.
So I'd say that a 40 Mhz mips 3000 should outperform any 486 in floating
point by quite a large margin. Certainly enough for me to want to
invest $700 in one.
BTW the bench numbers are: 4.94 mflops for the 3100 (16.67 mhz), 3.75
mflops for the 2100 (12.5 Mhz) and 3.70 mflops for the 486-66.