(1) For some reason, my computer intermittently processes only four projects at one time when it has capacity for five, Do I have a wrong setting?
(2) I noticed that processing priorities on various project seemed to be strange. I noticed that seti projects due February 19th were processed before Einstein projects due January 4,2009.

Thanks for your advise on these.

1/4/2010 8:06:53 PM Starting BOINC client version 6.10.24 for windows_x86_64
1/4/2010 8:06:53 PM log flags: file_xfer, sched_ops, task
1/4/2010 8:06:53 PM Libraries: libcurl/7.19.7 OpenSSL/0.9.8l zlib/1.2.3
1/4/2010 8:06:53 PM Data directory: D:\ProgramData\BOINC
1/4/2010 8:06:53 PM Running under account StaceyBaird
1/4/2010 8:06:54 PM Processor: 4 GenuineIntel Intel(R) Core(TM)2 Quad CPU Q8200 @ 2.33GHz [Intel64 Family 6 Model 23 Stepping 7]
1/4/2010 8:06:54 PM Processor: 2.00 MB cache
1/4/2010 8:06:54 PM Processor features: fpu tsc pae nx sse sse2 pni
1/4/2010 8:06:54 PM OS: Microsoft Windows 7: Home Premium x64 Edition, (06.01.7600.00)
1/4/2010 8:06:54 PM Memory: 4.00 GB physical, 8.00 GB virtual
1/4/2010 8:06:54 PM Disk: 298.09 GB total, 238.11 GB free
1/4/2010 8:06:54 PM Local time is UTC +8 hours
1/4/2010 8:06:54 PM NVIDIA GPU 0: GeForce 8400 GS (driver version 19562, CUDA version 3000, compute capability 1.1, 512MB, 29 GFLOPS peak)
1/4/2010 8:06:54 PM SETI@home Found app_info.xml; using anonymous platform
1/4/2010 8:06:54 PM rosetta@home URL http://boinc.bakerlab.org/rosetta/; Computer ID 1205399; resource share 100
1/4/2010 8:06:54 PM Docking@Home URL http://docking.cis.udel.edu/; Computer ID 53657; resource share 100
1/4/2010 8:06:54 PM Einstein@Home URL http://einstein.phys.uwm.edu/; Computer ID 2235920; resource share 100
1/4/2010 8:06:54 PM SETI@home URL http://setiathome.berkeley.edu/; Computer ID 5229037; resource share 100
1/4/2010 8:06:54 PM Einstein@Home General prefs: from Einstein@Home (last modified 01-Jan-2010 10:49:58)
1/4/2010 8:06:54 PM Einstein@Home Computer location: home
1/4/2010 8:06:54 PM Einstein@Home General prefs: no separate prefs for home; using your defaults
1/4/2010 8:06:54 PM Reading preferences override file
1/4/2010 8:06:54 PM Preferences:
1/4/2010 8:06:54 PM max memory usage when active: 2006.58MB
1/4/2010 8:06:54 PM max memory usage when idle: 2252.28MB
1/4/2010 8:06:54 PM max disk usage: 1.00GB
1/4/2010 8:06:54 PM (to change, visit the web site of an attached project,
1/4/2010 8:06:54 PM or click on Preferences)
1/4/2010 8:06:54 PM Using proxy info from GUI
1/4/2010 8:06:54 PM Not using a proxy

---

The welfare of the people is the highest law - Cicero

If no one complains, the people must be satisfied.

(1) There could be a number of explanations. Make sure you have GPU processing enabled in BOINC. Go to BOINC > Advanced > Preferences > and make sure "Use GPU while computer is in use" is ticked.
Also, if that doesn't work, go to Your Account > SETI@home preferences > Edit Preferences > and tick "Use NVIDIA GPU if present".

For number (2), I believe this has something to do with 'project debt', and you may be repaying your Einstein 'debt', since you have crunched SETI@home for more than it's allocated resource share. However, I am not sure on that one.

Hope it helps...

- Luke.

---

- Luke.

When you have only 4 task running is this 4 CPU & 0 GPU task? If that is true. Then you might see a message like this:
[Collatz Conjecture] Reporting 1 completed tasks, requesting new tasks for GPU
[Collatz Conjecture] Scheduler request completed: got 0 new tasks

I think a good option for a future version of BOINC would be to have project CPU prefs. To allow user to say I want to give S@H 4 cpus, rosetta 2 cpus, & folding 2 cpus. Perhaps something along those lines, but then I know things would become extremely complex for trying to schedule tasks.

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

Thank for the ideas.

what happens is a GPU project is one of the four projects that are being crunched. Thus it appears to me that one of the CPU's is not being used. The GPU is set to run at all times, although I do give it a rest sometimes if I need additional power.

I don't see any mention of the [collatz conjecture] in my Messages listings.

---

The welfare of the people is the highest law - Cicero

If no one complains, the people must be satisfied.

If you're using Einstein and have its options set to GPU crunch, then it sounds like your CPU and GPU are being used on one WU. This happened to me earlier on. There's a lot of postings on Einstein's message boards about how inefficient their CUDA app is. For me, I've turned off GPU crunch for Einstein, but have it enabled for SETI. I get a mix of Einstein and SETI WU's for the CPU (also a quad core), mostly Einstein, and SETI only for GPU crunch. I'm getting the most bang for my buck doing this. Everything is 100% busy all the time, except when of course a project goes down.

---

Thank for the ideas.

what happens is a GPU project is one of the four projects that are being crunched. Thus it appears to me that one of the CPU's is not being used. The GPU is set to run at all times, although I do give it a rest sometimes if I need additional power.

I don't see any mention of the [collatz conjecture] in my Messages listings.

Sorry for the confusing "[collatz conjecture]" is one of the projects I run. I was using as an example. I was thinking maybe you ran out of GPU work, liek many projects do.

With 1 GPU & 3 CPU task running do you have any task that are in "ready to start" status? Also if you open task manager do you see 100% cpu usage when this is occurring?

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

Yes, I do have Einstein tasks waiting to start. Right now it all looks like this:
Einstein ABP1CUDA Running (1.00 cpu + 1.00 nvidia)
Docking Charm Waiting
Docking Charm Running
Einstein 63.01 Waiting
Rosetta Mini Running
Rosetta Mini Running

All the CPU's are in use in tightly bunched waves up to 100 %

Upon looking at the above, it appears that one task, the ABP1CUDA, is using 100 percent of both one cpu and one gpu. That is the answer suggested by M Dawson below in a reply earlier. I think he must be correct.

Thanks to all for getting me straight on that.


With 1 GPU & 3 CPU task running do you have any task that are in "ready to start" status? Also if you open task manager do you see 100% cpu usage when this is occurring?

---

The welfare of the people is the highest law - Cicero

If no one complains, the people must be satisfied.

Yes, I do have Einstein tasks waiting to start. Right now it all looks like this:
Einstein ABP1CUDA Running (1.00 cpu + 1.00 nvidia)

Suggestion: Don't run Einstein GPU tasks (at least not yet).

Einstein's CUDA app is, IMHO, not worth running. It uses not only the whole GPU, but an entire CPU core also. Because of this, your system can only run three additional CPU tasks, for a total of four including Einstein.

Worse yet, Einstein barely uses the GPU at all; depending on your CPU/GPU combination you'll only see a very small increase in speed as compared to the CPU-only version of Einstein. Some people have even reported that it runs slower on their GPU! What's happening is that most of the work appears to be done on the CPU anyway, and you're tying up your GPU with very little benefit to show for it.

My strong recommendation is to disable Einstein GPU tasks on their preferences web page and do GPU crunching for a different project. Considering how fast your CPU is and how slow your graphics card is, I suspect that your Einstein tasks will run FASTER without the GPU. Then go ahead and run something else on your GPU if you want.

To put things in perspective, I have a vastly faster GPU than you do, and even with that monster card, I only saw a 33% improvement in execution speed on Einstein. Typically, on other projects with both CPU and GPU versions of the same application, I see around a 2000% increase in speed running on the GPU. The Einstein GPU app, from a performance standpoint, is simply awful.

---

Want to find one of the largest known primes? Try PrimeGrid. Or help cure disease at WCG.