| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Help Conquer Cancer
Thread: Low Work for Help Conquer Cancer ( Nov 20, 2012 ) |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 161
|
|
| Author |
|
|
kateiacy
Veteran Cruncher USA Joined: Jan 23, 2010 Post Count: 1027 Status: Offline Project Badges: 
|
For those having Nvidia boards do not forget that you can switch to GpuGrid who is Nvidia only. You can have both projects on the same machine and when the GPU is idle on WCG crunch on GpuGrid WUs. And for those with AMD/ATI cards, POEM can fulfill the same role. It runs on both AMD/ATI and Nvidia GPUs.  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Not to misunderstand, the backup function is external to WCG. The internal is the "If there is no work for...", which so far does not allow a secondary choice. You get whatever is available from the slots in shared memory.
Whilst, if using an app_info [advanced crunchers only] and specifying which CPU project to participate in addition to HCC on GPU (See the workaround thread), then all threads would run HFCC on the CPU, though not sure about those "reserved" CPU threads that are set to support GPU crunching. Think you loose those to CPU crunching, but the experts can and surely will correct me on that. |
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: 
|
Quick update,
Still no estimate on arrival of the new work units. We are still working towards getting more work units to you as soon as we can and will try to keep everyone updated as the status changes. As of right now, no status change, only work going out will be reruns. Thank you for your patience, -Uplinger |
||
|
|
B2I
Senior Cruncher usa Joined: Jan 23, 2011 Post Count: 232 Status: Offline Project Badges: 
|
I've noticed that after the HCC GPU tasks went dry and I pulled my app_info file, several other projects are reporting "No Tasks are available" : SNTS, GFAM, and DSFL. are they running low also?
----------------------------------------No worries, the other projects are keeping all my cores busy. Also, My two NVida machines that were not using an App_info file are still burning through a backlog of HCC. Should finish the last of them today.  |
||
|
|
RT
Master Cruncher USA - Texas - DFW Joined: Dec 22, 2004 Post Count: 2636 Status: Offline Project Badges: 
|
I am pretty much a WCG only kind of a guy. I have a computer that has an AMD GPU in it and one that has a NVIDIA GPU. Both have been running HCC GPU exclusively. This morning the AMD machine had drained the queue. The NVIDIA card machine is about to drain it's queue.
----------------------------------------So I am changing them to a device profile that runs CPU projects only and HFCC (I am of course renaming the app_info.xml )and restarting BOINC on those two machines. Now, I want to be among the first to know when the HCC WUs come available again so I can reverse the process and get back to cranking out the HCC GPU WUs on the GPU enabled machines and HCC CPU WUs on the CPU only machines. Can we get someone to send us an email when the WUs become available again for the various queues? (NVIDIA GPU WUs, AMD GPU WUs and CPU WUs)?? Thanks Much  |
||
|
|
Ingleside
Veteran Cruncher Norway Joined: Nov 19, 2005 Post Count: 974 Status: Offline Project Badges:
|
Whilst, if using an app_info [advanced crunchers only] and specifying which CPU project to participate in addition to HCC on GPU (See the workaround thread), then all threads would run HFCC on the CPU, though not sure about those "reserved" CPU threads that are set to support GPU crunching. Think you loose those to CPU crunching, but the experts can and surely will correct me on that. An application without any tasks won't use any resources either. Only if a task is running, or waiting on enough resources becoming available, will an application have any effects. If example you're running 4 GPU-tasks for HCC but 2 for another project, you can run 3 HCC at once and have GPU partially idle waiting until drops to 2 HCC before the other GPU-task starts. This can happen even if there's more HCC avaiable to run.  "I make so many mistakes. But then just think of all the mistakes I don't make, although I might." |
||
|
|
B2I
Senior Cruncher usa Joined: Jan 23, 2011 Post Count: 232 Status: Offline Project Badges:
|
Here's how I am coping. Just pull the App file, set your projects to HCC and make sure the GPU option and "...send work from another project" is checked . Check your system(s) often. When/if you see a single channel of gpu pick up, slap that app file back in. (don't forget to abort all the non HCC tasks first)
---------------------------------------- |
||
|
|
rodgers82
Cruncher Joined: Jul 13, 2009 Post Count: 7 Status: Offline Project Badges: 
|
I've noticed that after the HCC GPU tasks went dry and I pulled my app_info file, several other projects are reporting "No Tasks are available" : SNTS, GFAM, and DSFL. are they running low also? No worries, the other projects are keeping all my cores busy. Also, My two NVida machines that were not using an App_info file are still burning through a backlog of HCC. Should finish the last of them today. I've had the same issue. I remove WCG from boinc and reattached it. It started to download more work the moment I did that. |
||
|
|
Jim1348
Veteran Cruncher USA Joined: Jul 13, 2009 Post Count: 1066 Status: Offline Project Badges:
|
And for those with AMD/ATI cards, POEM can fulfill the same role. It runs on both AMD/ATI and Nvidia GPUs. I forgot to mention above that to get POEM to run as a backup when no HCC are available, you have to set "Allow research to run on my CPU?" to "No". Otherwise, a CPU job will run instead when the core is free. You probably can achieve the same thing by setting BOINC to run with less than the maximum number of processors (e.g., 90%), but I have not tried that yet. It may depend on your version of BOINC also. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I just saw the notice from Uplinger after seeing an error from BOINC saying I had no work. I did an update but could only get one WU. I am going to update my projects list to work on other projects.
---------------------------------------- Oops! I just went back to BOINC to review progress and I now have 7 WUs, but I will need to keep an eye on this to make sure I don't run out of work for all of my systems. If this appears to be happening, I will still need to add more projects back into my list. [Edit 1 times, last edit by Former Member at Nov 21, 2012 4:51:53 PM] |
||
|
|
|