Credit modification proposal -- comments?
Message boards :
Questions/Problems/Bugs :
Credit modification proposal -- comments?
Message board moderation
| Author | Message |
|---|---|
|
Send message Joined: 26 Jun 08 Posts: 645 Credit: 472,402,268 RAC: 255,774        |
Looking at the NFS@Home column of the project credit comparisons, it does not appear inappropriate to double the current amount of credit awarded for a completed work unit. In addition, while I will continue to factor in the credit claimed by the client as the number of calculations in the work units does vary, I propose to narrow the allowed credit range to within 20% of the credit of a "reference" work unit. Comments on this proposal? |
|
Send message Joined: 23 Sep 09 Posts: 3 Credit: 1,734,906 RAC: 0 
|
> it does not appear inappropriate Math/Law? :) I have no objections. Semi-OT - are there any plans to offer a GPU client for this project? If so then kidnapping Cluster Physik from Collatz/Milkyway would be a good first step. |
|
Send message Joined: 26 Jun 08 Posts: 645 Credit: 472,402,268 RAC: 255,774 |
Unfortunately, not anytime soon. I'm aware of one attempt to do lattice sieving in CUDA, and it was significantly slower than the CPU version we are currently using. |
|
Send message Joined: 26 Jun 08 Posts: 645 Credit: 472,402,268 RAC: 255,774 |
The changes to the granted credit have been made. I'm currently using a multiplier of 2, but that may be tweaked over time. |
|
Send message Joined: 10 Sep 09 Posts: 1 Credit: 951,415 RAC: 839    
|
The changes to the granted credit have been made. I'm currently using a multiplier of 2, but that may be tweaked over time. Thank you for adjusting the granted credit. It seems more 'fair' now compared to the other BOINC projects. BlackGoose05 
|
|
Send message Joined: 8 Sep 09 Posts: 2 Credit: 4,314 RAC: 0 |
Thank you :) |
|
Send message Joined: 7 Sep 09 Posts: 3 Credit: 1,600,775 RAC: 238    
|
Thank you for adjusting the granted credit. It seems more 'fair' now compared to the other BOINC projects. Agreed. Thank you! |
|
Send message Joined: 19 Sep 09 Posts: 1 Credit: 1,072,523 RAC: 0
|
Nice! I'll be back once my monthly run of SIMAP is done. Happy crunching! |
|
Send message Joined: 5 Oct 09 Posts: 3 Credit: 46,541,037 RAC: 0   
|
Looking at the NFS@Home column of the project credit comparisons, it does not appear inappropriate to double the current amount of credit awarded for a completed work unit. In addition, while I will continue to factor in the credit claimed by the client as the number of calculations in the work units does vary, I propose to narrow the allowed credit range to within 20% of the credit of a "reference" work unit. Comments on this proposal? Why not just set a fixed credit amount for the WUs here since they seem quite predictable? It looks like the cap is set at 28 for the current WUs so why not set that as the fixed amount? The BOINC benchmark system is totally broken so any credit calculations based on it are silly. Stop cheating before it starts, give equal credit for equal work: many other projects have recently moved to fixed, server based credit systems. There are a lot of users that will not run projects based on the BOINC benchmarks. Server based credits are easy to implement and no one can complain about them not being fair. Maybe set up a poll and let people vote. Docking did that and the vote was 50 something to 0 in favor of fixed or server based credits. Thanks for a really nice running project! Regards/Beyond |
