Found the problem Speedy. You added the segment after the app_info closing tag. I PMed you the way it should look. Just go up to just above where you added the segment and find </app_info> delete it there and enter it at the bottom of the new segment.

---

PROUD MEMBER OF Team Starfire World BOINC

This is looking like the ol' proverbial can o' worms to the kittyman.

Question....
Are these settings mostly GPU specific, or does the system CPU capability play into it too?

Perhaps somebody could start to compile a chart with recommended starting points for these settings based on the GPU.
There is going to be an onslaught of questions regarding this once the app leaves beta and is added to the Lunatics installer.

Yeah that's why I'm going to hold off on this till there are some more testing by the more willing. Once things have been 'nailed down' somewhat then I'll give this a try. Once my main rig become a backup rig I may attempt this on the 480 not sure about trying this on a 250. Just looks like a pain.

---

Traveling through space at ~67,000mph!

Thanks Perryjay. I'll test it when there's plenty of AP work ready to send.I'm currently working on 11 AP CPU tasks. These will finish tomorrow.

---

Someone may want to take a look at this one http://setiathome.berkeley.edu/result.php?resultid=2013361612.

A three second long run time and completed successfully would seem to indicate a problem for an AP.

Also, all text details below ": Considering profile 'compute_20' for gpu='sm_21' in 'cuModuleLoadDataEx_4' is missing from the posted result.

How did this happen? Near as I can tell I induced it when I was suspending all mb work to force the NV AP's (2 available at the time) to crunch through. I was doing that because the estimated run times are so huge when you are below the 10 Completed threshold that BOINC is not asking for more work, and I am quickly running out of MB's to crunch.

Lt

Someone may want to take a look at this one http://setiathome.berkeley.edu/result.php?resultid=2013361612.

A three second long run time and completed successfully would seem to indicate a problem for an AP.

Also, all text details below ": Considering profile 'compute_20' for gpu='sm_21' in 'cuModuleLoadDataEx_4' is missing from the posted result.

How did this happen? Near as I can tell I induced it when I was suspending all mb work to force the NV AP's (2 available at the time) to crunch through. I was doing that because the estimated run times are so huge when you are below the 10 Completed threshold that BOINC is not asking for more work, and I am quickly running out of MB's to crunch.

Lt

ap_02ap11ad_B6_P0_00391_20110726_01488.wu and
ap_02ap11ad_B6_P0_00390_20110726_01488.wu are probably similar, being only 13.42 seconds apart from the same channel. What's missing from the stderr is most likely the

In ap_remove_radar.cpp: get_indices_to_randomize: num_ffts_forecast < 100. Blanking too much RFI?

shown in Task 2013361586. It really means that all 2047 data chunks are near enough to a detected RFI spot to need blanking, so there's no sense in processing the WU.

                                                                Joe

Not a problem, your stderr.txt is just missing the following:

In ap_remove_radar.cpp: get_indices_to_randomize: num_ffts_forecast < 100. Blanking too much RFI?

a lot of the Wu's with B6_P0 in them end like this, as it happens i've just been clearing out some of my tasks with B6_P0 in them:

Valid Astropulse v505 tasks for computer 5427475

Claggy

Thanks Joe, Claggy!

No problems, Keep on Crunching!

Lt

This is the first time I've ever heard of AP tasks finishing so quickly.

---

Got it set up then, Speedy? I don't see any new GPU APs yet on your rig. I hope everything is working okay.

---

PROUD MEMBER OF Team Starfire World BOINC

Got it set up then, Speedy? I don't see any new GPU APs yet on your rig. I hope everything is working okay.

No I'm still not able to get tasks due to lack of ready to send. In my previous post I was saying it's the first time I've heard of tasks finishing in seconds.

---

Oh, yeah, those are the ones with 100% blanking. Can't crunch them if there's nothing there to crunch. Just like the -9 MB tasks with too much noise.

By the way, you can use Efmer's newest rescheduler to move CPU APs over to your GPU. I haven't tried it as I've still got enough but it's a thought.

---

PROUD MEMBER OF Team Starfire World BOINC

Now the results come in from my wingmates ... and my WUs get inconclusive :(
http://setiathome.berkeley.edu/results.php?hostid=5790778&offset=0&show_names=0&state=0&appid=5

Now the results come in from my wingmates ... and my WUs get inconclusive :(
http://setiathome.berkeley.edu/results.php?hostid=5790778&offset=0&show_names=0&state=0&appid=5

Hm... will see what third result gives...

Now the results come in from my wingmates ... and my WUs get inconclusive :(
http://setiathome.berkeley.edu/results.php?hostid=5790778&offset=0&show_names=0&state=0&appid=5

Hm... will see what third result gives...

Raistmer, he's getting the single pulses right but keeps getting 30 repetitive pulses on all his work. The only ones he's validated are 30/30s. (though it was nice to get over 1700 on a 30/30 :-) ) That's why everything else is going inconclusive.

---

PROUD MEMBER OF Team Starfire World BOINC

I see, let's await what third result give.

Now the results come in from my wingmates ... and my WUs get inconclusive :(
http://setiathome.berkeley.edu/results.php?hostid=5790778&offset=0&show_names=0&state=0&appid=5

You managed to catch very noisy bunch of tasks BTW :)

Now the results come in from my wingmates ... and my WUs get inconclusive :(
http://setiathome.berkeley.edu/results.php?hostid=5790778&offset=0&show_names=0&state=0&appid=5

First marked invalid

http://setiathome.berkeley.edu/workunit.php?wuid=788645897

---

With each crime and every kindness we birth our future.

Oh, yeah, those are the ones with 100% blanking. Can't crunch them if there's nothing there to crunch. Just like the -9 MB tasks with too much noise.

By the way, you can use Efmer's newest rescheduler to move CPU APs over to your GPU. I haven't tried it as I've still got enough but it's a thought.

Works quite well actually.

Edit to add: No problems with the beta at all on my end running dual GTX 560ti's. It's crunched probably 60 valid tasks with no invalids.

Great work guys.

---

Executive Director GPU Users Group Inc. -
brad@gpuug.org

Can someone please post where I can get Efmer's newest rescheduler tool?

---

Can someone please post where I can get Efmer's newest rescheduler tool?

http://www.efmer.eu/forum_tt/index.php?topic=428.0

---

Executive Director GPU Users Group Inc. -
brad@gpuug.org