Re: [Gems-users] Effect of OPAL_RUBY_MULTIPLIER on CPI

[Dan Gibson <degibson@xxxxxxxx>]

Art (for me, with my checkpoints) is rife with TLB misses. Opal's modeling of TLB fill is... dubious at best. 
Try a different benchmark and see if you get a more intuitive response.

On Tue, Mar 30, 2010 at 9:52 PM, <ubaid001@xxxxxxx> wrote:

1. Nope, no chnages. 2. Art Open MP benchmark. 3. Opal sim-step for 100 million instructions.
Suhail

On Mar 30 2010, Dan Gibson wrote:

I agree -- this behavior seems odd. As I've said before, I'm no Opal expert,
but I'll cover the basics before giving up:
1. Are you changing anything else? (Code or configuration?)
2. What workload are you running?
3. How are you deciding how long to run the simulation? sim-step
something-or-other?

Regards,
Dan

On Tue, Mar 30, 2010 at 8:43 PM, <ubaid001@xxxxxxx> wrote:

Hi,

While running simulations on OPAL + Ruby simulator , on increasing the
OPAL_RUBY_MULTIPLIER value, the CPI decreases.
Since we make the memory system slower with a higher value of the
multiplier, we expected an increase in the CPI. But that does not seem to be
the case. Can anyone shed more light on this?


_______________________________________________
Gems-users mailing list
Gems-users@xxxxxxxxxxx
https://lists.cs.wisc.edu/mailman/listinfo/gems-users
Use Google to search the GEMS Users mailing list by adding "site:
https://lists.cs.wisc.edu/archive/gems-users/" to your search.

_______________________________________________
Gems-users mailing list
Gems-users@xxxxxxxxxxx
https://lists.cs.wisc.edu/mailman/listinfo/gems-users
Use Google to search the GEMS Users mailing list by adding "site:https://lists.cs.wisc.edu/archive/gems-users/" to your search.

-- 
http://www.cs.wisc.edu/~gibson [esc]:wq!