Folding Forum

Hello friends

I'm not sure but i think this server is giving all my clients the same WU.

Have any of you ever experience this problem, or is this normal, my clients have been running for months with no problems.

Except for 59** 1888 point WU's.

Thanks for your help

Project 5913 uses FahCore_14. That particular FahCore is different than any other FahCore in several respects. The load that it puts on the GPU is significantly less that 100% and tends to be erratic. [There's adequate allowance for those charactistics in the benchmarking process.] Also, the project can use multiple results from the same WU, so assignments like that are not a problem.

Ok so let me get this clear.

So is ok to get the same WU on many clients at the same time, like i did.

Because the only time i seen this, is when i had the same client ID on a few GPU's, and they were all folding the same WU.

Thanks for your help.

If all 4 of those GPU clients are on the same computer then it could mean that you haven't given a unique MachineID to each client.

They are all in the same system, and they all have different ID's.

As long as you have unique MachineIDs, you'll get credit for all of them. If you have duplicate MachineIDs, you very likely will NOT get credit for all of them.

We do see rare cases of duplicates from other projects and the Pande Group would like to minimize that sort of duplication -- only one successful completion of a WU is scientifically useful. FahCore_14 is unique in that they're all useful so there is no need to minimize the duplication.

As Bruce has already mentioned, duplicates are possible with core 14 and ihaque has also stated the same:

viewtopic.php?f=52&t=13542&p=132285#p132285

Let us know if you receive the full 1888 for all 4 WUs.

I didn't get credit for the 1888 point WU's, if i remove the advmethods, will that stop my client from getting them?

They take a long time to finish, and who knows if all that work is lost,cause the server won't get them back.