[parsec-users] New inputs for benchmarks

Yungang Bao ybao at CS.Princeton.EDU
Tue Aug 23 21:45:37 EDT 2011


Hi Tom,

Most of the native inputs of PARSEC are from real world. And then the sim-xxxxx inputs are scaled down from the native inputs. Unfortunately we don't have other real world inputs. If you have larger inputs, you can scale them down by yourself to meet your requirements. That paper presents some insights on how to keep similar behavior while scaling down input sets.

Best Regards,
Yungang

----- Original Message -----
From: "Thomas Patson" <thomas.patson at gmail.com>
To: "PARSEC Users" <parsec-users at lists.cs.princeton.edu>
Sent: Tuesday, August 23, 2011 7:08:00 PM
Subject: Re: [parsec-users] New inputs for benchmarks


Thanks Yungang, however I have a follow-up question. I looked into that paper which you mentioned and I have a few questions 


Is it possible to get hold of the real world inputs that were scaled down ? I could not find a reference. Sorry if I missed it. 
I intend to scale down the real world input but keep the working sets even bigger than what they are for the native input, hence 
the need to get the real world inputs that were used , as mentioned in the paper 


Thanks again! 





On Mon, Aug 22, 2011 at 8:20 PM, Yungang Bao < ybao at cs.princeton.edu > wrote: 


Hi Tom, 

We are sorry we don't have scripts to generate larger inputs. But you can generate inputs by yourself. There is one paper on PARSEC inputs which may be useful: "Fidelity and Scaling of the PARSEC Benchmark Inputs" ( http://parsec.cs.princeton.edu/publications/bienia10inputs.pdf ) 

Best Regards, 
Yungang 




----- Original Message ----- 
From: "Thomas Patson" < thomas.patson at gmail.com > 
To: parsec-users at lists.cs.princeton.edu 
Sent: Monday, August 22, 2011 7:57:15 PM 
Subject: [parsec-users] New inputs for benchmarks 


Hi, 


I was wondering if I could get a pointer to generating inputs for the parsec benchmarks that have larger working sets ( even larger than the native inputs that have been provided ). Is there a document available or even some scripts ? Should I just look at the test and simsmall inputs and see what they are on a per-benchmark basis and then generate inputs for each benchmark ? Are there any scripts that can be helpful and save me some time ? Some sort of guidance would be very much appreciated. 


Thanks! 


Tom 
_______________________________________________ 
parsec-users mailing list 
parsec-users at lists.cs.princeton.edu 
https://lists.cs.princeton.edu/mailman/listinfo/parsec-users 
_______________________________________________ 
parsec-users mailing list 
parsec-users at lists.cs.princeton.edu 
https://lists.cs.princeton.edu/mailman/listinfo/parsec-users 


_______________________________________________
parsec-users mailing list
parsec-users at lists.cs.princeton.edu
https://lists.cs.princeton.edu/mailman/listinfo/parsec-users


More information about the parsec-users mailing list