The last NVIDIA WU that I received was on Sep. 27. Every time my machine requests GPU work (several times per day), I get the "no jobs available" message. Still getting CPU work OK (when I can connect...). I haven't seen this mentioned by anyone else, so is it just me?? Just for the record, my machine continues to receive and crunch GPU WU's from Einstein@Home.

---

What about now? Today I suddenly got a few WU for GPU.

---

Was in the same boat, but just got a few. Hang in there!

Edit: 4min after posting I got 22 cuda tasks:)

Same here, no GPU tasks.

I did order a second GTX460 last week, hope for next week there will be GPU work available to test my 2 X gtx460 with amd six-cores PC. Btw i had to flash the asus mainboard bios to newest version. Old version would not detect two nvidia cards. And, you dont need an SLI cable for boinc to use 2 cards, you just have to follow instructions (the xml file thing).

I was in the same condition, but received GPUs for the first time just 5 hours ago.

UPDATE: I have received 99 GPUs so far. More coming it! Someone gave the scheduler a kick!!!

Jeff

---

This morning I asked for some GPU work and expected to receive the same message I've been getting for the last 2 weeks (no work for GPU) But I got this!10/7/2011 4:54:40 AM | SETI@home | Finished download of libfftw3f-3-1-1a_upx.dll
10/7/2011 4:54:40 AM | SETI@home | Started download of seti_609.jpg
10/7/2011 4:54:43 AM | SETI@home | Finished download of seti_609.jpg
10/7/2011 4:54:43 AM | SETI@home | Started download of cudart32_30_14.dll
10/7/2011 4:54:54 AM | SETI@home | Finished download of cudart32_30_14.dll
10/7/2011 4:54:54 AM | SETI@home | Started download of cufft32_30_14.dll
10/7/2011 4:56:09 AM | SETI@home | Finished download of setiathome_6.10_windows_intelx86__cuda_fermi.exe
10/7/2011 4:56:09 AM | SETI@home | Started download of setiathome-6.10_cuda_AUTHORS
10/7/2011 4:56:39 AM | SETI@home | Finished download of setiathome-6.10_cuda_AUTHORS
10/7/2011 4:56:39 AM | SETI@home | Started download of setiathome-6.10_cuda_COPYING
10/7/2011 4:57:04 AM | SETI@home | Finished download of setiathome-6.10_cuda_COPYING
10/7/2011 4:57:04 AM | SETI@home | Started download of setiathome-6.10_cuda_COPYRIGHT
10/7/2011 4:57:05 AM | SETI@home | Finished download of setiathome-6.10_cuda_COPYRIGHT
10/7/2011 4:57:05 AM | SETI@home | Started download of setiathome-6.10_cuda_README
10/7/2011 4:57:21 AM | SETI@home | Finished download of setiathome-6.10_cuda_README
And after this downloaded I received several GPU WU. They were shorties but WU none the less. They also have unusually long names such as 02ap11ad.6741.22562.6.6.10.40_0 I don't recall seeing these # before. Have I been upgraded, downgraded, set aside or what? The WU keep coming in and I'll keep crunchen. Just wondered what has happened!!!! Pick

---

This morning I asked for some GPU work and expected to receive the same message I've been getting for the last 2 weeks (no work for GPU) But I got this!

<snip>

And after this downloaded I received several GPU WU. They were shorties but WU none the less. They also have unusually long names such as 02ap11ad.6741.22562.6.6.10.40_0 I don't recall seeing these # before. Have I been upgraded, downgraded, set aside or what? The WU keep coming in and I'll keep crunchen. Just wondered what has happened!!!! Pick

What happened was that the BOINC/SETI admin team fixed a server bug which had been silently introduced sometime after 14 September, and which prevented users of the ordinary Berkeley CUDA application getting any new work.

I say 'silently', because on 13 September they'd introduced a completely different bug which had seriously affected users of optimised applications. Everybody was so busy watching and worrying about that bug that nobody noticed the stock bug, until they got the servers running properly on Wednesday this week. (Most people on this message board run the optimised applications, and spend their time talking about that).

What you saw was, first, the downloading of the various components of the application needed to process WUs on your GPU, and then the workunits. The task names are absolutely normal - they've been in that format ever since we've had access to the 'ALFA' multi-beam antenna at Arecibo, more years ago than I care to remember.

Thanks for the info, and so quick. Much appreciated, Pick

---

Hm should Astropulse ATI lunactrics(?) Apps work now too?
Because i installed it now on a server with a low end 4350 (BOINC Soft recognizing it) and get only the no work message (app_info seems to work got no error about it). So i dont know if it is now an config error or still a Seti error. Profil is set to accept other applications.

<app_info>
<app>
<name>astropulse_v505</name>
</app>
<file_info>
<name>ap_5.06_win_x86_SSE2_OpenCL_ATI_r521.exe</name>
<executable/>
</file_info>
<file_info>
<name>AstroPulse_Kernels_r521.cl</name>
<executable/>
</file_info>
<app_version>
<app_name>astropulse_v505</app_name>
<version_num>505</version_num>
<platform>windows_intelx86</platform>
<avg_ncpus>0.04</avg_ncpus>
<max_ncpus>0.20</max_ncpus>
<plan_class>cuda</plan_class>
<cmdline>-instances_per_device 1 -hp -unroll 10</cmdline>
<flops>30987654321</flops>
<file_ref>
<file_name>ap_5.06_win_x86_SSE2_OpenCL_ATI_r521.exe</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>AstroPulse_Kernels_r521.cl</file_name>
<copy_file/>
</file_ref>
<coproc>
<type>ATI</type>
<count>1</count>
</coproc>
</app_version>
</app_info>

---

Hm should Astropulse ATI lunactrics(?) Apps work now too?
Because i installed it now on a server with a low end 4350 (BOINC Soft recognizing it) and get only the no work message (app_info seems to work got no error about it). So i dont know if it is now an config error or still a Seti error. Profil is set to accept other applications.

ATI work fetch does work with Astropuse_v505 there is no problem there, but you have to patient as AP Wu's are quite rare,

Your app_info on the other hand shouldn't work, it has both ATI and cuda in it, best to run the Lunatics Installer and let that put in the right app_info,

Claggy

He already got 2 AP units.
The first exited with too much RFI.

Change plan class in your appinfo to "ati13ati"

---

With each crime and every kindness we birth our future.

He already got 2 AP units.
The first exited with too much RFI.

Change plan class in your appinfo to "ati13ati"

Yes, that's the conventional value - but it makes no difference at all in an app_info. Getting it right really only matters when installing an app_info file on a computer which already has a stock of work.

Don't change it if you have work cached and assigned to the existing plan class.

AHH all works :) thx *GGG*

Edit: Unfortune it "kills" the whole machine (fileserver) so absolute no performance is left, had to stop it :(

---

AHH all works :) thx *GGG*

Edit: Unfortune it "kills" the whole machine (fileserver) so absolute no performance is left, had to stop it :(

Then put in some reduced commandline parameter's so host is usable, (and don't use the High Priority switch), like:

<cmdline>-instances_per_device 1 -unroll 6 -ffa_block 4096 -ffa_block_fetch 2048</cmdline>

Claggy

Is there something I need to do to make sure I get gpu units with stock app? I set the preferences on the account page to cpu - no and nvidia gpu - yes, it wouldn't give out work. If I set both to yes I would get some cpu work.

Should I just try the lunatics installer 0.38 and see if it gets any units for either gpu (fermi and a cayman) or is this still a lack of workunits to send out thing?

Thanks.

---

Is there something I need to do to make sure I get gpu units with stock app? I set the preferences on the account page to cpu - no and nvidia gpu - yes, it wouldn't give out work. If I set both to yes I would get some cpu work.

Should I just try the lunatics installer 0.38 and see if it gets any units for either gpu (fermi and a cayman) or is this still a lack of workunits to send out thing?

Thanks.

It should be possible to get work with the "cpu - no and nvidia gpu - yes" settings, but everything is very busy and running slow - it may take a long time to catch any.

Ah ok, thanks for the help. I guess I'll just set both cpu and gpu allowed now and just let it sort itself out.

---

The last NVIDIA WU that I received was on Sep. 27. Every time my machine requests GPU work (several times per day), I get the "no jobs available" message. Still getting CPU work OK (when I can connect...). I haven't seen this mentioned by anyone else, so is it just me??

I finally got one GPU job early this morning ... no more since. But I guess things are still catching up.

---

AHH all works :) thx *GGG*

Edit: Unfortune it "kills" the whole machine (fileserver) so absolute no performance is left, had to stop it :(

Then put in some reduced commandline parameter's so host is usable, (and don't use the High Priority switch), like:

<cmdline>-instances_per_device 1 -unroll 6 -ffa_block 4096 -ffa_block_fetch 2048</cmdline>

Claggy

Ah thx works now @ full Network speed again, only little Video chunknees every ~10 Minutes, so i changed BOINC Config to compute only between 0:30 and 20:30 so i have the TV primetime free for lagfree MediaPC Connectivity.

---