Hey GE ZHIGUO,

I have tried to run parsec same benchmark with same input
(blackscholes with simlarge input - in_64K.txt)through command line
and measure it num of instruction by perf tool.

I am find that number of instruction executed is in range on 5000 x
10^6. while with GEM5 running parsec gives the num of instruction as
85 x 10^6 and execution time is around 15 min.

so, i am feeling some thing wrong on how i am executing the parsec with GEM5.

if you or someone in USER group know how to get correct results to
run parsec with X86 pl. guide.
I have used following
1. disk image on http://www.cs.utexas.edu/~parsec_m5/
2. instructions as per
http://gem5user.blogspot.com/2013/02/running-parsec-benchmarks-on-gem5.html

--
A.K.T

On Fri, Feb 21, 2014 at 2:32 AM, GE ZHIGUO <[email protected]> wrote:
> Hi, Tewar
> Some PARSEC benchmarks produce output files or prints.
> You can diff the simulation outputs with those obtained from the direct 
> executions on top of desktop computers.
>
> Thanks!
>
> Zhiguo
>
> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On 
> Behalf Of AKT
> Sent: Friday, February 21, 2014 5:23 AM
> To: [email protected]
> Subject: Re: [gem5-users] GEM5 run with PARSEC blackscholes - Seems to be 
> early termination
>
> Hello Udit, GE ZHIGUO,
>
> I am seeing same type of output with PARSEC while running and it is
> completed in 15 min. same as Udit shown in the earlier email.
>
> Can you guide me some thing how to know what results i am getting is
> correct?
>
> Hope you can guide me
>
> Regards
> A.K.Tewar
>
>
>
> _______________________________________________
> 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
_______________________________________________
gem5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users

Reply via email to