HP3000-L Archives

May 1995, Week 2

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Stan Sieler <[log in to unmask]>
Reply To:
Stan Sieler <[log in to unmask]>
Date:
Thu, 11 May 1995 14:57:58 -0700
Content-Type:
text/plain
Parts/Attachments:
text/plain (25 lines)
Wirt writes:
>
> If the 917, 927, 937, 947, 918 series processors are taken to be 10 on the
> performance scale, then the 9x9KS series are as follows:
 
...
but, 917 and 918 are measurably different.
 
   CPU  Clock                   Sieve benchmark
   ---  ----------------------- ---------------
   917  32 MHz                       276,410          100%
   918  34 MHz + NO superscalar      256,837          107.6%
   918  34 MHz + superscalar         233,860          118.1%
 
   Note: 34 vs 32 MHz --> 1.065%
 
(Normally, a 918 is always in superscalar mode)
 
My timings of the two machines show the 918 about 18% faster.
(This is strictly CPU hog timings, with no OS calls of any kind, and no
floating point usage.)
 
Note that the 918 (without superscalar) vs. 917 is about what would be
predicted from the ratio of their respective clock speeds.

ATOM RSS1 RSS2