Re: [Gems-users] g_endpoint_bandwidth have negative impact on performance?

["Min Xu (Hsu)" <xu@xxxxxxxxxxx>]

Could it be the workload variability? If the slowdown is not
big, it may well be it.

On Thu, 20 Oct 2005 Liqun Cheng wrote :
> Hi there,
> 
> I have increased the value of g_endpoint_bandwidth, from 10000 to 30000. but
> the half benchmarks take longer to finish. Is there any possility that
> increased bandwidth screws up the adaptive routing algorithm?
> 
> thanks
> Legion

> _______________________________________________
> Gems-users mailing list
> Gems-users@xxxxxxxxxxx
> https://lists.cs.wisc.edu/mailman/listinfo/gems-users