| Author |
Message |
|---|
 Michael Goetz Volunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14038
ID: 53948
Credit: 479,665,689
RAC: 431,032
                               
|
|
New work has been loaded into the GFN524288 port in PRPNET, picking up at b=700K (where we stopped in BOINC).
____________
My lucky number is 75898524288+1 |
|
|
|
|
New work has been loaded into the GFN524288 port in PRPNET, picking up at b=700K (where we stopped in BOINC).
I don't see it in the master ini file. Could you post the server address and port?
____________
|
|
|
|
|
New work has been loaded into the GFN524288 port in PRPNET, picking up at b=700K (where we stopped in BOINC).
I don't see it in the master ini file. Could you post the server address and port?
prpnet.mine.nu:11001
____________
There's someone in our head but it's not us. |
|
|
|
|
|
Thank you sir!
____________
|
|
|
|
|
New work has been loaded into the GFN524288 port in PRPNET, picking up at b=700K (where we stopped in BOINC).
Is this for CPU or GPU or both?
I ask because I'm trying to get it to run on GPU and having issues.
____________
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14038
ID: 53948
Credit: 479,665,689
RAC: 431,032
|
New work has been loaded into the GFN524288 port in PRPNET, picking up at b=700K (where we stopped in BOINC).
Is this for CPU or GPU or both?
I ask because I'm trying to get it to run on GPU and having issues.
On PRPNet you can run it on anything you want, provided it works. :)
At the current B level, we've got about 100K to go on GeneferCUDA, but only about 25K or so on GenefX64. So you can run on the GPU for a while, but on a CPU there's only a little ways to go before the slower Genefer80 will need to be used.
It should run on the same GPUs it ran on in BOINC, i.e., any CC 1.3 or higher CUDA GPU.
What kind of issues are you having?
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
I'm trying to run it on a evga gtx 470 and I get this error:
./genefercuda: error while loading shared libraries: libcufft.so.3: wrong ELF class: ELFCLASS64
Unable to determine version from './genefercuda'
EDIT: I just figured it out. I was using this line
export LD_LIBRARY_PATH=/usr/local/cuda/lib64
Now using this line and it worked
export LD_LIBRARY_PATH=/usr/local/cuda/lib64:/usr/local/cuda/lib
____________
|
|
|
|
|
|
I'm trying to run 2 cards in the same box on this project.
It seem to only run on 1 card when I fire off the 2-dual-start-prpclient.sh.
Do I set something in the GPU affinity line?
____________
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14038
ID: 53948
Credit: 479,665,689
RAC: 431,032
|
I'm trying to run 2 cards in the same box on this project.
It seem to only run on 1 card when I fire off the 2-dual-start-prpclient.sh.
Do I set something in the GPU affinity line?
Um... Yeah. I built in a feature to permit doing exactly what you need to do. Let me go look up how it works (it's been a year) and I'll post the procedure, if nobody else beats me to it...
____________
My lucky number is 75898524288+1 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14038
ID: 53948
Credit: 479,665,689
RAC: 431,032
|
|
It's set in prpclient.ini:
// This sets the GPU affinity for CUDA apps on multi-GPU machines. It defaults to
// -1, which means that the CUDA app can run on an GPU.
gpuaffinity=
0 is the first GPU, 1 is the second GPU, etc.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
That did it. Thank you!
____________
|
|
|
