That sounds pretty high to me, but it'd be helpful if you provided some
additional details: are you starting simulation from a region of interest
or are you starting from the shell using ./start_script? how many cores are
you simulating? Is the workload multi-threaded?


On Wed, Jul 31, 2013 at 7:01 PM, Mansour Shafaei Moghaddam <
[email protected]> wrote:

> Hi all,
>
> I have got a log from the memory requests arrived to the memory controller
> and the results reveal that the majority of requests (around 80% in my
> benchmark) are initiated by kernel. I have two questions regarding that and
> I really appreciate if you could share your opinions about them:
>
> 1) Does this number sound rational or not?
> 2) Are all of the benchmark's memory requests included in the non-Kernel
> requests? Can we say that all of the non-Kernel requests are exclusively
> originated from the benchmark?
>
> P.S. I know that this percentage may be different depending on the
> application I run.
>
> Thanks in advance
>
> _______________________________________________
> http://www.marss86.org
> Marss86-Devel mailing list
> [email protected]
> https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel
>
>
_______________________________________________
http://www.marss86.org
Marss86-Devel mailing list
[email protected]
https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel

Reply via email to