Message boards :
Wish list :
Top GPU models
Message board moderation
| Author | Message |
|---|---|
DamaralandSend message Joined: 7 Nov 09 Posts: 152 Credit: 16,181,924 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
I found out this on POEM. I think it's usefull to have some figures on every GPU/OS: http://boinc.fzk.de/poem/gpu_list.php Whis this project would have something like this... if possible. Think should't be too complicated with all the database info. |
skgivenSend message Joined: 23 Apr 09 Posts: 3968 Credit: 1,995,359,260 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Already exists, http://www.gpugrid.net/gpu_list.php Should also be useful, http://www.gpugrid.net/forum_thread.php?id=2507 FAQ's HOW TO: - Opt out of Beta Tests - Ask for Help |
DamaralandSend message Joined: 7 Nov 09 Posts: 152 Credit: 16,181,924 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
:( I'm sorry I didn't saw it |
DamaralandSend message Joined: 7 Nov 09 Posts: 152 Credit: 16,181,924 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Is it possible that it ranks? on Linux(1.000) GeForce GTX 275 on Windows (0.492) GeForce GTX 275 Is it due to be an average of people's cards? or due to the drivers on both systems? If it's an average and standard deviation, maybe interesenting know how many of each because may be affected by overclockers |
skgivenSend message Joined: 23 Apr 09 Posts: 3968 Credit: 1,995,359,260 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
The Windows table is a reasonable fair reflection of GPU performance, though you would need to double the rank of the GTX590, as it has 2 GPU's. The Linux list looks like a mess. Don't go by it. I don't know how the list is generated but I guess short tasks vs long and different credit systems might have impacted on this significantly. It may also be the case that only a couple of GTX275's were used and only for long tasks, while most people ran a mix, some has bad settings, and the few top Fermi users mostly or exclusively run short tasks. Perhaps down-clocking has played a part too. FAQ's HOW TO: - Opt out of Beta Tests - Ask for Help |
©2025 Universitat Pompeu Fabra