[RC5] Pentium 4 Prescott Processors, and the ROT function.

Joseph Kaye jkaye at isd.net
Tue Oct 12 15:32:12 EDT 2004


Thanks Mark,

  It looks like things are improving in the Intel camp quite a bit - 
nice to see.

-Joseph


Mark Duell wrote:

> I cant get this to send to the list for some reason, so here you go:
>
> 2.8Ghz Prescott with HT on:
> dnetc v2.9008-490-GTR-04050719 for Win32 (WindowsNT 5.1).
> [Oct 12 17:31:56 UTC] RC5-72: using core #0 (SES 1-pipe).
> [Oct 12 17:32:16 UTC] RC5-72: Benchmark for core #0 (SES 1-pipe)
>                      0.00:00:16.23 [4,591,365 keys/sec]
> [Oct 12 17:32:16 UTC] RC5-72: using core #1 (SES 2-pipe).
> [Oct 12 17:32:34 UTC] RC5-72: Benchmark for core #1 (SES 2-pipe)
>                      0.00:00:16.15 [5,791,752 keys/sec]
> [Oct 12 17:32:34 UTC] RC5-72: using core #2 (DG 2-pipe).
> [Oct 12 17:32:55 UTC] RC5-72: Benchmark for core #2 (DG 2-pipe)
>                      0.00:00:17.48 [5,703,615 keys/sec]
> [Oct 12 17:32:55 UTC] RC5-72: using core #3 (DG 3-pipe).
> [Oct 12 17:33:14 UTC] RC5-72: Benchmark for core #3 (DG 3-pipe)
>                      0.00:00:16.18 [5,550,096 keys/sec]
> [Oct 12 17:33:14 UTC] RC5-72: using core #4 (DG 3-pipe alt).
> [Oct 12 17:33:33 UTC] RC5-72: Benchmark for core #4 (DG 3-pipe alt)
>                      0.00:00:16.23 [5,745,785 keys/sec]
> [Oct 12 17:33:33 UTC] RC5-72: using core #5 (SS 2-pipe).
> [Oct 12 17:33:52 UTC] RC5-72: Benchmark for core #5 (SS 2-pipe)
>                      0.00:00:16.15 [5,613,337 keys/sec]
> [Oct 12 17:33:52 UTC] RC5-72: using core #6 (GO 2-pipe).
> [Oct 12 17:34:11 UTC] RC5-72: Benchmark for core #6 (GO 2-pipe)
>                      0.00:00:16.10 [6,344,712 keys/sec]
> [Oct 12 17:34:11 UTC] RC5-72: using core #7 (SGP 3-pipe).
> [Oct 12 17:34:30 UTC] RC5-72: Benchmark for core #7 (SGP 3-pipe)
>                      0.00:00:16.12 [6,447,878 keys/sec]
>
> Compare to 
> http://n0cgi.distributed.net/speed/query.php?cputype=79&cpumhz=2800&recordid=1&contest=rc572&multi=0 
>
>
> Mark Duell
>
> Joseph Kaye wrote:
>
>> Has anyone benchched one of these yet?    Now that the rotate 
>> function is back in the processor, I azm curious about it's 
>> performance.   I think the uber long pipeline will still hinder it's 
>> performance.
>>
>> -Joseph
>> _______________________________________________
>> rc5 mailing list
>> rc5 at lists.distributed.net
>> http://lists.distributed.net/mailman/listinfo/rc5
>>
>
>



More information about the rc5 mailing list