home *** CD-ROM | disk | FTP | other *** search
/ NetNews Usenet Archive 1992 #31 / NN_1992_31.iso / spool / comp / benchmar / 1899 < prev    next >
Encoding:
Text File  |  1992-12-30  |  1.5 KB  |  39 lines
  1. Newsgroups: comp.benchmarks
  2. Path: sparky!uunet!zaphod.mps.ohio-state.edu!rpi!batcomputer!vger.cheme.cornell.edu!OLIN
  3. From: olin@vger.cheme.cornell.edu (Steve Thompson)
  4. Subject: Re: Alpha, RISC/6000, HP benchmarks
  5. Message-ID: <1992Dec30.114222.4840@tc.cornell.edu>
  6. Sender: news@tc.cornell.edu
  7. Nntp-Posting-Host: vger.cheme.cornell.edu
  8. Reply-To: olin@vger.cheme.cornell.edu
  9. Organization: Chemical Engineering, Cornell University, NY
  10. References: <1992Dec29.204830.12667@zooid.guild.org>,<1992Dec30.113719.4722@tc.cornell.edu>
  11. Date: Wed, 30 Dec 1992 11:42:22 GMT
  12. Lines: 25
  13.  
  14. I wrote:
  15.  
  16. >Having just received a DEC 3000/400-AXP, I have had time to run one
  17. >benchmark. This is written in C, and solves the lorenz equations
  18. >using a Runge-Kutta 4 algorithm. Here are the CPU times, using the
  19. >highest optimization level in each case, on some machines that we
  20. >have:
  21. >
  22. >IBM RS6000-530H, AIX 3.2, XLF 2.2:        13.29 seconds
  23. >VAX 4000-500, VMS 5.5-1, FORT 5.8:         9.85 seconds
  24. >HP 9000/720, HPUX 8.07:             7.55 seconds
  25. >DEC 3000/400, OpenVMS AXP 1.0, DEC C 1.1:     2.79 seconds
  26. >
  27.  
  28. The compiler for the RS6000 should have been XLC 1.2, not XLF 2.2. I
  29. can also add a time for the RS6000-350: 10.74 seconds.
  30.  
  31. Steve
  32.  
  33. ---------------------------------------------------------------------------
  34. Steve Thompson, System Mangler   Internet: thompson@cheme.cornell.edu
  35. School of Chemical Engineering   Bitnet:   thompson@crnlchme
  36. Olin Hall, Cornell University    Phone:    (607) 255 5573
  37. Ithaca NY 14853
  38. ---------------------------------------------------------------------------
  39.