ATI Card requesting work . . . how to stop

rwremote
rwremote
Joined: 2 Apr 11
Posts: 3
Credit: 1003202
RAC: 0
Topic 195811

Have 2 PC's new to crunching EH.
Boinc Client 6.10.60 on one and 6.10.58 on the other. Both clients have Do NOT use GPU checked (or unchecked can't remember - not in front of the PC's now). And EH preferences on the Web set for do not use Nvidia GPU.

But both of these PC's keep requesting GPU work :-

2011-05-31 23:51:13.5592 [PID=11611]    [send] CPU: req 0.00 sec, 0.00 instances; est delay 0.00
2011-05-31 23:51:13.5592 [PID=11611]    [send] ATI: req 51840.00 sec, 1.00 instances; est delay 0.00


and

2011-05-31 23:51:13.5592 [PID=11611]    [send] CPU: req 0.00 sec, 0.00 instances; est delay 0.00
2011-05-31 23:51:13.5592 [PID=11611]    [send] ATI: req 51840.00 sec, 1.00 instances; est delay 0.00
.
.
.
.
2011-05-31 23:51:15.6540 [PID=11611] [debug]   [HOST#4135286] MSG(high) No work sent
2011-05-31 23:51:15.6540 [PID=11611] [debug]   [HOST#4135286] MSG(high) Your computer has no NVIDIA GPU
2011-05-31 23:51:15.6540 [PID=11611]    Sending reply to [HOST#4135286]: 0 results, delay req 60.00
2011-05-31 23:51:15.6543 [PID=11611]    Scheduler ran 2.112 seconds

These started off being repeated every 5 minutes but then backed off to every 30 minutes and look to be about every hour now.
Is there a setting I've missed or will this right itself in due course ?

Cheers
Rob

Gundolf Jahn
Gundolf Jahn
Joined: 1 Mar 05
Posts: 1079
Credit: 341280
RAC: 0

ATI Card requesting work . . . how to stop

The only thing you've missed is the "NVIDIA" in the preference label of "Use NVIDIA GPU no" ;-)

Since Einstein@home doesn't have any application for ATi GPUs, it also lacks the corresponding preference.

So, yes, it will "right" itself in due course, in that the frequence of the requests will be reduced to once in 24 hours.

Gruß,
Gundolf

Computer sind nicht alles im Leben. (Kleiner Scherz)

rwremote
rwremote
Joined: 2 Apr 11
Posts: 3
Credit: 1003202
RAC: 0

Thanks Gundolf, Its fixed

Thanks Gundolf,

Its fixed itself - nothing done from this side.

2011-06-02 03:29:11.1874 [PID=8965 ]    [send] CPU: req 6041.98 sec, 1.00 instances; est delay 0.00
2011-06-02 03:29:11.1874 [PID=8965 ]    [send] ATI: req 0.00 sec, 0.00 instances; est delay 0.00
2011-06-02 03:29:11.1874 [PID=8965 ]    [send] work_req_seconds: 6041.98 secs
and
2011-06-02 04:25:07.2791 [PID=19772]    [send] CPU: req 82.23 sec, 0.00 instances; est delay 0.00
2011-06-02 04:25:07.2791 [PID=19772]    [send] ATI: req 0.00 sec, 0.00 instances; est delay 0.00
.
.
2011-06-02 04:25:07.2858 [PID=19772]    [version] Host lacks CUDA coprocessor for plan class (BRP3cuda32)
.
.
.
2011-06-02 03:29:13.2692 [PID=8965 ]    [send] don't need more work
2011-06-02 03:29:13.2705 [PID=8965 ]    Sending reply to [HOST#4135287]: 1 results, delay req 60.00
2011-06-02 03:29:13.2708 [PID=8965 ]    Scheduler ran 2.091 seconds

Should really have just left it alone and not posted the first message :)

Cheers

robertmiles
robertmiles
Joined: 8 Oct 09
Posts: 127
Credit: 21888035
RAC: 32017

I've found some messages on

I've found some messages on various BOINC projects say that the current versions of BOINC don't offer any way to fully stop them from asking for workunits of all three types (CPU, Nvidia GPU, and AMD/ATI GPU), if the computer has those resources. However, BOINC projects with sufficiently recent versions of the BOINC server software can send back a response say wait for up to nearly a week before asking that BOINC project for any workunits of that type again.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.