[parsec-users] characterization of parsec 2.0

Yungang Bao ybao at CS.Princeton.EDU
Mon Oct 17 17:07:25 EDT 2011


Hi Sparsh,

The working set data in Table 2.1 should be obtained by running simulation with simlarge inputs. 
Working sets that are ‘unbounded’ are large and have the additional qualitative property that there is significant application demand to make them even bigger.

Best Regards,
Yungang

----- Original Message -----
From: "sparsh mittal" <sparsh0mittal at gmail.com>
To: "PARSEC Users" <parsec-users at lists.cs.princeton.edu>
Sent: Monday, October 17, 2011 4:48:55 PM
Subject: Re: [parsec-users] characterization of parsec 2.0


Thanks. Table 2.1 helps a lot and serves the purpose. 
I want to seek clarification. 
1. It is not clear whether Table 2.1 is for simlarge or native. I assume simlarge. 

2. 
Table 2.1 states, ferret WS as unbounded and facesim WS as large. 
Table 4.1 states, ferret WS2 as 64MB, and facesim WS2 as 256MB. 

The text says, Programs with unbounded working sets are canneal, dedup, ferret and freqmine. Does it refer to native? 

Actually all I need to do is to classify Parsec into three groups: unbounded, large and small(+medium). And I am unclear of ferret and facesim. 


Can you clarify. 
Thanks and Regards 
Sparsh Mittal 





On Mon, Oct 17, 2011 at 3:29 PM, Yungang Bao < ybao at cs.princeton.edu > wrote: 


Hi Sparsh, 

Then there is no such characterization data for raytrace yet. You can find some other characterizations in Table 2.1 and Table 
2.2. Hope they are helpful for you. 


Best Regards, 
Yungang 

----- Original Message ----- 
From: "sparsh mittal" < sparsh0mittal at gmail.com > 



To: "PARSEC Users" < parsec-users at lists.cs.princeton.edu > 
Sent: Monday, October 17, 2011 3:55:27 PM 
Subject: Re: [parsec-users] characterization of parsec 2.0 


Actually Table 4.1 in the thesis lists characteristic for only 12 benchmark, which are also given in research paper The PARSEC Benchmark Suite: Characterization and Architectural Implications. Thus raytrace is still not there. 


On Mon, Oct 17, 2011 at 2:19 PM, sparsh mittal < sparsh0mittal at gmail.com > wrote: 


Thanks 



Thanks and Regards 
Sparsh Mittal 








On Mon, Oct 17, 2011 at 2:06 PM, Yungang Bao < ybao at cs.princeton.edu > wrote: 


Hi Sparsh, 

Raytrace was introduced in PARSEC 2.0, and you can find its characterization in Dr. Bienia's phd thesis. 
http://parsec.cs.princeton.edu/publications/bienia11benchmarking.pdf 

Best Regards, 
Yungang 




----- Original Message ----- 
From: "sparsh mittal" < sparsh0mittal at gmail.com > 
To: parsec-users at lists.cs.princeton.edu 
Sent: Monday, October 17, 2011 12:47:40 PM 
Subject: [parsec-users] characterization of parsec 2.0 


Hello 
I saw the paper by Bienia which characterizes parsec-1. Is there a paper describing characterization of parsec 2.0 ? I need characteristics of raytrace also. 
If behavior has not changed much, just let me know. I am using simlarge. 
Thanks and Regards 
Sparsh Mittal 



_______________________________________________ 
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 
_______________________________________________ 
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