Date: | Mon, 31 May 2010 13:26:04 -0500 |
---|---|
From: | Byn Choi <bynchoi1@xxxxxxxxxxxx> |
Subject: | [Gems-users] SPLASH-2 barnes input |
Hello-I'd like to conduct a little survey. I'm trying to run SPLASH-2 barnes in a simics+GEMS environment, and I've been able to run up to 2048 bodies with a reasonable simulation time (2.5 hrs). Bigger inputs require longer simulation time as well as much larger overall memory footprint. This is still rather small compared to the input size used in the original SPLASH-2 paper (The SPLASH-2 Programs: Characterization and Methodological Considerations (ISCA'95)). So I'm wondering, how big of an input have you guys been able to use successfully? Byn |
[← Prev in Thread] | Current Thread | [Next in Thread→] |
---|---|---|
|
Previous by Date: | Re: [Gems-users] traffic in Garnet, Niket Agarwal |
---|---|
Next by Date: | , (nil) |
Previous by Thread: | Re: [Gems-users] Specifying OPAL to simulate only till the program ends, Dan Gibson |
Next by Thread: | [Gems-users] suitable protocol, sparsh mittal ISU |
Indexes: | [Date] [Thread] |