Hi!

My nVidia 9600GSO equipped PC only runs for 4-6 hours a day. That is enough to complete a WU within 2-3 days. The problem is, the BOINC scheduler typically queues 3 more work units for me to process while the first is being worked on. This typically results in the loss of three work units due to deadline expiration.

What can be done to get me a more conservative scheduling? Is there some setting that I can make in my BOINC client?

By the way I prefer GPUGrid over Folding@Home because it has less impact on CPU usage and it keeps the graphics of the system much snappier, compared to F@H. Good job guys!


Christian

I now know what causes it to download four work units. It's the fact that I have four CPU cores (Intel Q6600).

I am still using the .10 beta version of the client because the .11, .12 and .13 versions reportedly have been duds. I will try the suggested fix above to see if that helps.

If you change it to zero new WUs then, in theory, it will not download any new ones. However, that doesn't actually fix your problem overall. If you only want 1 new WU at a time....then try 0.25 and see if that works...


HTH

I now know what causes it to download four work units. It's the fact that I have four CPU cores (Intel Q6600).

I am still using the .10 beta version of the client because the .11, .12 and .13 versions reportedly have been duds. I will try the suggested fix above to see if that helps.

Try upgrading to 6.3.14, It is more stable than even 6.3.10 and has a better behaving scheduler. I've tried hard in four days to break it and can't. Rom released versions today on the web download page for both Windows and Linux and released it for all alpha testers to test. 6.3.14 is the first 6.3.XX version he has asked all alpha testers to start testing on. Note it is still labeled as development version.

If you change it to zero new WUs then, in theory, it will not download any new ones. However, that doesn't actually fix your problem overall. If you only want 1 new WU at a time....then try 0.25 and see if that works...


HTH

Just wanted to add: I tried the above myself on all 3 rigs. It works somewhat with WCG, but not with GpuGrid....the rigs (quads) still maintain 4 WUs.

I have suggested we need a GPU limit, like the CPU limit, but they need to be two different items, currently they are the same but only a CPU limit. It will be considered on how to do and may be in a future versions. This will avoid two problems, not getting enough work for GPUs when you have less CPUs than GPUs or getting to much work extra work when you have more CPUs than GPUs. In essence you should end up with always 1 running and 1 on standby for each GPU where you have enough CPUs to support them. The current way was jsut an extension of the exisiting CPU limits, which we have proved to be not sufficient.