Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!newstop!sun!amdcad!dvorak.amd.com!proton!tim
From: tim@proton.amd.com (Tim Olson)
Newsgroups: comp.benchmarks
Subject: Re: Byte benchmark of a Sun 2/120
Message-ID: <1991Mar12.165237.12624@dvorak.amd.com>
Date: 12 Mar 91 16:52:37 GMT
References: <8816@plains.NoDak.edu>
Sender: usenet@dvorak.amd.com (Usenet News)
Reply-To: tim@amd.com (Tim Olson)
Organization: Advanced Micro Devices, Austin, TX
Lines: 22

In article <8816@plains.NoDak.edu> swiers@plains.NoDak.edu (Mike Swiers ) writes:

| =============================================================================
| Arithmetic Test (type = register): 10000 Iterations
| -----------------------------------------------------------------------------
|                        Arithmetric        Geometric         Variance
|                               Mean             Mean         (6 tests)
| User Time (secs):            52.58            52.58              0.00
| System Time (secs):           0.10             0.10              0.00
| Real Time (secs):            52.80            52.80              0.00

Be aware that the Arithmetic Test in the Byte benchmark simply
performs 1 operation each of +, -, *, and / in a loop, weighing each
operation type equally.  In "typical" code, however, they should be
weighted more like 55% add, 34% sub, 10% mul, 1% div. Also, since
integer multiply and divide are usually an order of magnitude slower
than addition or subtraction, this benchmark ends up mainly reporting
the performance of these slower, less-used operations.
--
	-- Tim Olson
	Advanced Micro Devices
	(tim@amd.com)