Of course Christina is correct, if we wanted to get serious about these benchmarks a number of replications would be required. I wouldn't want to be on John's system for this.

I ran a second benchmark and included some additional information about the system status. I suspect the first run was quite similar to this one, maybe excel might have been open as well, but no concurrent requests on my part during either test. Although an N of 2 is twice as good as 1, it is a far cry from strong evidence, but I suspect the 11 minute 40 second time interval would prove to be pretty stable upon multiple replications under OS/2. I'm not so sure under Win NT and even less certain than the Windows 95 estimates would be stable. Time. OS Chip Clock Speed RAM SAS hh:mm:ss:fs ------------------ ----------- ------------ --- ---- ----------- OS/2 with fixpak 8 Pentium II 400 Mhz 384 6.12 11:44:27 OS/2 with fixpak 8 Pentium II 400 Mhz 384 6.12 11:40:46* OS/2 with fixpak 8 Pentium II 400 Mhz 384 7.00 7:55:85

Windows NT Pentium III 500 Mhz 384 6.12 x:xx:xx Windows NT Pentium III 500 Mhz 384 7.00 6:08:54

Windows 95 Pentium I? 133 Mhz 48 6.12 1:57:49:78

x currently unavailable

* WinOS/2 session with Office 4.3 open with Word and Eudora open FileStar/2 open, Powerchute open, Networking Software enabled but none of these processes being actively exploited during the test

___________________________________________________________

Kevin F. Spratt, Ph.D. Iowa Testing Programs & Spine Diagnostic & Treatment Center & Iowa Spine Research Center 224-D Lindquist Center University of Iowa Iowa City, Iowa 52242 (319) 335-5572 (voice) (319) 335-6399 (fax) Kevin-Spratt@Uiowa.edu (e-mail)