further NEC2 benchmark results

        I found that the NEC2 source code that I was compiling earlier
was of unknown parentage and had been customized by some kind soul.
The "modifications" slowed it down something fierce.

        In my previous message I posted some results from running some
pre-compiled code that had been used for some other benchmark results that
were posted to the list. The numbers looked pretty good, however I have
recompiled the code to take advantage of our platform. Here are the
optimized results:

   Platform Precision Processors Fill Factor Run
______________________________________________________________________
Sun UltraSparc1 Single 1 16.168 69.313 88.042
   @167MHz Single 2 16.165 39.608 58.301

        " Double 1 22.208 66.927 92.155
        " Double 2 21.280 37.304 61.958

        compiled with SunPro f77 compiler version 3.0.2
        used -fast -parallel command line options

        Ray Anderson
        raymonda_at_radium.eng.sun.com
Received on Tue May 21 1996 - 20:38:00 EDT