Gould PN9000 Performance

[Mike Muuss]

Ron Natalie recently stated that a Gould PN 9000 machine could
deliver 10 x VAX780 for $400K CPU price, and the fellow from DEC
asked "do you really believe that?".

About 8 weeks ago, I published the results of my benchmarks on the
Gould machines, which can be sumarized simply:

32/8750 (1 CPU):	claim 5xVAX780, measured 4.9xVAX780
32/8780 (2 CPUs):	claim 10xVAX780, measured 9.8xVAX780

>From what I know of Gould's products, these relationships hold:
	PN 9050 is >= 32/8750
	PN 9080 is >= 32/8780

I know of two other UNIX tests done (by others) on PN9050 machines,
one indicated 4.5 x VAX780, the other indicated 5.5 x VAX780.

I really believe this, because I ran it myself on a machine which
was provided to me (at my request) with blank disks and their
distribution.  I loaded their UNIX in, and ran the tests;  no
Gould staff were permitted to access the machine while the tests
were running.  Unless there where hallucinogenic gasses in the
machine room during the tests, I have confidence in my tests.

Anybody wishing the full details of my tests can send me mail.

As with all benchmarks, remember that "your mileage may vary".

Best,
 -Mike Muuss

(301)-278-6678
  AV  283-6678
  FTS 939-6678

ArpaNet:  Mike @ BRL
UUCP:     ...!{decvax,cbosgd,seismo}!brl-bmd!mike
Postal:
  Mike Muuss
  Leader, Advanced Computer Systems Team
  Computer Techniques and Analysis Branch
  Systems Engineering and Concepts Analysis Division
  U.S. Army Ballistic Research Laboratory
  Attn: AMXBR-SECAD (Muuss)
  APG, MD  21005