If you're in syscall emulation mode, each workload is treated completely
independently.  So for your example of 2X and 5Y benchmarks, you're really
running 7 independent benchmarks from gem5's point of view. Unfortunately,
this leads to some inaccuracy, since things like code that would be shared
across the different instances in a real machine will not be shared in the
simulator.

Steve



On Mon, Mar 24, 2014 at 9:35 AM, Otoom, Mof <[email protected]> wrote:

> Hi,
>>
>> I'm an Assistant Professor of Computer Engineering, interested in
>> Computer Architecture. I should first thank you so much for the great
>> effort in developing and maintaining the great simulator, GEM5.
>>
>> I'm new in using GEM5, and I need to do a simple experiment using GEM5. I
>> wish you can help me.
>>
>> In my experiment I need to run different number of instances of two
>> benchmarks. Let's say I have X and Y benchmarks and I need to run 2X and 5Y
>> on two multicore architecture, so that I can evaluate the performance of
>> these architectures. The two architectures can be differentiated by the
>> number of CPUs, types of ISAs, execution order, memory hierarchy, etc.
>>
>>
>> I really would appreciate any directions on that.
>>
>> Thanks!
>>
>>
>> Regards,
>> Mof Otoom, Ph.D.
>>
>
>
> _______________________________________________
> gem5-users mailing list
> [email protected]
> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
>
_______________________________________________
gem5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users

Reply via email to