[Beowulf] scaling SPECint_base2000 for harpertown

[Tom Elken]

> [mailto:beowulf-bounces at beowulf.org] On Behalf Of andrew holway

> I recently got SPECint_base2000 figures from intel for 
> harpertown using low optimisation flags (-O2 -pthread -fPIC). 
> This benchmark is commonly used in particle physics circles 
> to provide a common base.

> I was expecting the numbers to scale pretty much exactly but 
> if you look at the link below the faster the chip the more 
> SPECint_base2000 you get per GHz.
> 
> Is this expected? These are aggregate figures for a dual 
> socket machine with 1.5GB ram per core.
> 
> http://www.moonet.co.uk/2004

>From what I can interpret, this URL shows SPECint_base2000 = 11751
for a 3.0 GHz Harpertown at -O2 optimization.

The highest result on the SPEC website for SPECint_base2000 I could find
is 3108 with full optimization (-fast +FDO, etc.).  3.0 GHz Harpertown
is almost surely faster than that "Precision Workstation 390 (Intel Core
2 Extreme processor X6800, 2.93 GHz)" that achieved the 3108, but not by
a factor greater than 3.

How did you come up with the 11751 score?

If you were running SPECint_rate_base2000 at 8 cores, at full
optimization, the score would likely be in the 200-300 range.

Cheers,
-Tom Elken


> 
> Cheers
> 
> Andy
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org To change your 
> subscription (digest mode or unsubscribe) visit 
> http://www.beowulf.org/mailman/listinfo/beowulf
>