Finally got my HD 7950 to work with BOINC (had 32 bit W10 installed instead of 64). CPU is a FX 6300.

Currently I'm using my CPU for Rosetta@Home (5 cores, 1 saved for GPU).

1. Is there any way to run more WUs at one time? Currently crunching 42k gflop units in about 80 seconds. Wonder if 2 WUs simultaneously would be more efficient?

2. Any settings I can change to make my GPU run more efficiently? Down clocking? Under volting?

3. If I do run another WU, would I also need to use another CPU core?

Thanks in advance

Finally got my HD 7950 to work with BOINC (had 32 bit W10 installed instead of 64). CPU is a FX 6300.

Currently I'm using my CPU for Rosetta@Home (5 cores, 1 saved for GPU).

1. Is there any way to run more WUs at one time? Currently crunching 42k gflop units in about 80 seconds. Wonder if 2 WUs simultaneously would be more efficient?

2. Any settings I can change to make my GPU run more efficiently? Down clocking? Under volting?

3. If I do run another WU, would I also need to use another CPU core?

Thanks in advance

MilkyWay is pretty efficient with their app, do you use GPU-Z or MSIAfterburner to monitor your gpu? If not get one of those and see what percentage of the gpu is being used per workunit, mine is 99% most of the time meaning no there is no room for a 2nd workunit to run at the same time.

Yup, MSI Afterburner.

It runs at 99% during the WU, so I guess that is it then. Oh well, figured I'd ask. Not sure why I'm not getting any new tasks though. I did 285 yesterday but haven't had any for quite a bit (not a full 24 hours).

Yup, MSI Afterburner.

It runs at 99% during the WU, so I guess that is it then. Oh well, figured I'd ask. Not sure why I'm not getting any new tasks though. I did 285 yesterday but haven't had any for quite a bit (not a full 24 hours).

Do a manual update and you should get around 300 new tasks. MW has a mandatory 10 minute backoff before getting new gpu tasks from the last time the pc asked for new tasks. So if you keep asking it won't send any, what most people do is setup a zero resource share gpu project to crunch a workunit or two while MW rests and then sends another 300, their max per gpu at one time.