[parsec-users] Looking for Windows Users

Christian Bienia cbienia at CS.Princeton.EDU
Mon Jan 18 12:21:27 EST 2010

Hi Jim,


Great to hear about your work. Many PARSEC workloads should already build
and run under Windows because Windows support was part of the original
development work. The PARSEC framework doesn't make this feature accessible
via build configurations but it's possible to use it manually.


We are generally happy to accept any code contributions that might be useful
for other PARSEC users. If you have anything more specific please let me
know and we can talk about the details.








From: parsec-users-bounces at lists.cs.princeton.edu
[mailto:parsec-users-bounces at lists.cs.princeton.edu] On Behalf Of Jim
Sent: Saturday, January 16, 2010 11:48 AM
To: parsec-users at lists.cs.princeton.edu
Subject: [parsec-users] Looking for Windows Users




I am a new user to PARSEC.


First, I wish to thank all those who contributed to this collection of
benchmark programs.


At this time I am focused on Windows platforms, at some future date this
will be expanded to Linux platforms. If any of you others following this
mailing list are using Windows I would like to co-ordinate our experiences
using similarly formatted thread subject lines.


I represent a small software development company that has produced a new
threading tool (www.quickthreadprogramming.com). The current version of the
tool is Windows-only. And when time and budget permit I intend to port to


After I adapt the benchmark programs to include a variation for my threading
tool, I wish to return a copy of the modified code back to Princeton for
inclusion into a subsequent release of PARSEC (assuming they approve of my
submissions). Regardless as to if you have my threading tool, the
programming strategy I use to optimize the benchmark may be of interest to
users of the other threading tools. Being relatively new with PARSEC I am
unsure of the procedure for resubmission of code. I would appreciate any
suggestions as to how to go about return of modified code.


Currently I have adapted the BodyTrack application for use with this new
threading tool. (with pleasing results)


There may be some issues involved in return of modified code that will have
to be worked out.


The first of which is my development environment is MS Windows with MS
Visual Studio (2005) together with Intel Parallel Studio. Your archives have
the files listed as .C type of files. My threading tool requires C++ and the
modifications will likely use C++0x. I took the liberty of renaming the
source files to .CPP for use in my projects. And I am making Visual Studio
Solution and Project files. This may not align with your archiving
requirements. I would appreciate advice from anyone with similar experiences
as to how to go about resubmission of code like this back to Princeton.


The second problem is my threading tool is a commercial product (principally
template headers and .LIB files). The additional code I will add to the
PARSEC apps will be conditionally compiled (as are your other threading
tools, e.g. TBB). While my modifications to the PARSEC apps will be public
domain (or under whatever license the main.c for the app has), my run time
library (and headers) will require licensing (the license you would use for
non-commercial purposes is currently a token $10). The conditional
modifications to the PARSEC apps will include a link as to where to obtain
the threading tool as well as where to obtain the most recent updates to the
PARSEC benchmarks that may have not yet been incorporated back into the
official PARSEC release).


Additionally, I am interested in locating Windows sites that currently have
PARSEC and would be willing to run the modified benchmark programs on their
systems (using their copy of the input data files). My collection of systems
does not include higher-end newer systems. Occasionally I do get access to
other systems. Next week I will have access to a Dell R610 w/ 2x Nehalem EX
5570. Potentially later the same site will have a Dell R900 4x Xeon 7400 (6
core each no HT). We all know that the PARSEC benchmark programs have large
input test data files. This makes it unsuitable for using a free service
such as Intel's Parallel Universe Portal. I would appreciate anyone offering
temporary use of their system.


Sincerely yours,


Jim Dempsey

QuickThread Programming, LLC



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cs.princeton.edu/pipermail/parsec-users/attachments/20100118/5a7dae0b/attachment.htm>

More information about the parsec-users mailing list