Message boards :
Number crunching :
Credits
Message board moderation
Previous · 1 · 2
| Author | Message |
|---|---|
skgivenSend message Joined: 23 Apr 09 Posts: 3968 Credit: 1,995,359,260 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
It's not pretty and still needs some work but the present opt-out system facilitates a massive gulf in credits between some projects. This fundamentally undermines the concept of a credit system. I think it's been discussed elsewhere and over a long time but for GPU projects it would need to be based on app complexity, utilization, power usage, GFlops and take into account the relative performance of a similar app on a CPU. Fortunately all of these things can be measured with the exception of complexity which could be estimated and agreed upon. There are lots of reasons to crunch and credits generally isn't too high on the list, but the worst reason now is probably project wide Boinc credits because it isn't uniform and therefore misrepresents contribution. FAQ's HOW TO: - Opt out of Beta Tests - Ask for Help |
|
Send message Joined: 2 Jan 09 Posts: 303 Credit: 7,321,800,090 RAC: 245 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
It's not pretty and still needs some work but the present opt-out system facilitates a massive gulf in credits between some projects. This fundamentally undermines the concept of a credit system. I think it's been discussed elsewhere and over a long time but for GPU projects it would need to be based on app complexity, utilization, power usage, GFlops and take into account the relative performance of a similar app on a CPU. Fortunately all of these things can be measured with the exception of complexity which could be estimated and agreed upon. There are lots of reasons to crunch and credits generally isn't too high on the list, but the worst reason now is probably project wide Boinc credits because it isn't uniform and therefore misrepresents contribution. AGREED, the current credit system only works for each project, not when comparing credits between projects. |
|
Send message Joined: 5 Mar 13 Posts: 348 Credit: 0 RAC: 0 Level ![]() Scientific publications ![]() |
It's tricky enough keeping credits consistent inside this project. I don't even want to imagine the mess in comparing to others. |
skgivenSend message Joined: 23 Apr 09 Posts: 3968 Credit: 1,995,359,260 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
It's tricky enough keeping credits consistent inside this project. I don't even want to imagine the mess in comparing to others. Don't even concern yourself with such issues, they are out of your hands. Focus on your research ;) FAQ's HOW TO: - Opt out of Beta Tests - Ask for Help |
|
Send message Joined: 19 Aug 07 Posts: 46 Credit: 45,339,082 RAC: 34 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]()
|
Credit / WU is calculated based on runtime of a WU on one of our machines. So we run a WU on one of our GPUs check time required to run and multiply by credit/time. Then there is also the bonus for the long tasks and also the fact that we usually round the credits up. Additionally we sometimes submit the WU's with the wrong credits out of a small mistake in a input file (but we usually catch those). In this same system used for working out how much credit is given to the CPU work unit? |
©2025 Universitat Pompeu Fabra