the faulty tasks seem to be back (erroring out after a few seconds):

http://www.gpugrid.net/result.php?resultid=21331546

:-(

I had a task fail after few seconds.

Stderr says: ERROR: file pme.cpp line 91: PME NX too small

here the URL: http://www.gpugrid.net/result.php?resultid=21429528

anyone any idea what was going wrong?

At least it went wrong for everyone, not just for you. A bad workunit.

WU 16799014

here another one, from this morning, with error message:

ERROR: file mdioload.cpp line 81: Unable to read bincoordfile

http://www.gpugrid.net/result.php?resultid=21431713

Same here
http://www.gpugrid.net/result.php?resultid=21432948
http://www.gpugrid.net/result.php?resultid=21432946
http://www.gpugrid.net/result.php?resultid=21431340
http://www.gpugrid.net/result.php?resultid=21431266
http://www.gpugrid.net/result.php?resultid=21430771
...and more others, all CUDA 80

Same here
http://www.gpugrid.net/result.php?resultid=21432948
http://www.gpugrid.net/result.php?resultid=21432946
http://www.gpugrid.net/result.php?resultid=21431340
http://www.gpugrid.net/result.php?resultid=21431266
http://www.gpugrid.net/result.php?resultid=21430771
...and more others, all CUDA 80

Until the new app (ACEMD3) is released, you should assign this host to a venue which receives work only from the ACEMD3 queue, as the other two queues have the old client, which is incompatible with the Turing cards.

obviously, the faulty tasks are back, here the next one from a minute ago:
http://www.gpugrid.net/result.php?resultid=21433016

This is even worse in times where new tasks are very rare, anyway :-(

the next ones:

http://www.gpugrid.net/result.php?resultid=21462742

http://www.gpugrid.net/result.php?resultid=21462460

http://www.gpugrid.net/result.php?resultid=21462682

http://www.gpugrid.net/result.php?resultid=21462715

they all didn't run even one second :-(

and here some more:

http://www.gpugrid.net/result.php?resultid=21463119

http://www.gpugrid.net/result.php?resultid=21463047

http://www.gpugrid.net/result.php?resultid=21463010

http://www.gpugrid.net/result.php?resultid=21462974

http://www.gpugrid.net/result.php?resultid=21463183

http://www.gpugrid.net/result.php?resultid=21463207

I think the license of the v9.22 app has expired this time.

I think the license of the v9.22 app has expired this time.

that's what I now am suspecting, too :-(

Any prediction when continous supply of new WU's will become available again ?
Nearly full month of very intermittent and small numbers of WU's.

Einstein is a happy project in the meantime :-)

Are all efforts being put into support of the new 20XX cards at the detriment
of the current 10XX cards ? (limited staff available maybe/lack of funding ?)

this is an increasingly annoying situation:

while there are no tasks available most of the time, some of the few ones that are being downloaded fail after 5 seconds:

http://www.gpugrid.net/result.php?resultid=21481323

ERROR: file mdioload.cpp line 81: Unable to read bincoordfile

:-( :-( :-(

Hi:

I see this is a well used section of the forum.

I would like to contribute some useful results here with my Alienware laptop but I have a high failure rate which I would like to resolve here. The GPU in my laptop is a Geoforce 660M. The OS I am using is uptodate Windows 10.

I would appreciate it if a tech person could narrow down the reason or reasons why I am experiencing such a high failure rate.

Clive Hunt
Canada

I would like to contribute some useful results here with my Alienware laptop

I am afraid that laptop GPUs are not made for this kind of load :-(

I would like to contribute some useful results here with my Alienware laptop

I am afraid that laptop GPUs are not made for this kind of heavy load :-(

I would like to contribute some useful results here with my Alienware laptop

I am afraid that laptop GPUs are not made for this kind of heavy load :-(

My Dell G7 15 laptop is happily crunching. That is another matter that I have to send a blast of air every day to get the dust-out.

Hi:

I see this is a well used section of the forum.

I would like to contribute some useful results here with my Alienware laptop but I have a high failure rate which I would like to resolve here. The GPU in my laptop is a Geoforce 660M. The OS I am using is uptodate Windows 10.

I would appreciate it if a tech person could narrow down the reason or reasons why I am experiencing such a high failure rate.

Clive Hunt
Canada

The issue is with the Scheduler on the GPUgrid servers. The Scheduler is sending CUDA65 tasks to your Laptop, all of which will fail due to an expired license. (Server end)
Your laptop can process CUDA80 tasks, but you are at the mercy of the Scheduler. For most Hosts it sends the correct tasks, and for a handful of Hosts, it is sending the wrong tasks.
This issue tends to affect Kepler GPUs (600 series GPU), even though they are still supported.
Some relevant posts discussing this issue are here:
http://www.gpugrid.net/forum_thread.php?id=5000&nowrap=true#52924
http://www.gpugrid.net/forum_thread.php?id=5000&nowrap=true#52920

The Project is in the middle of changing the Application to a newer version, hopefully when the new Application is released (ACEMD3), these issues will be smoothed out.

... when the new Application is released (ACEMD3)...

I am curious WHEN this will be the case