Please set <cpu_sched_debug> and send message log

Martin wrote:
> This has just got to be "non-optimal"...
> 
> I run s...@h at a low priority amongst other projects that have a higher
> priority.
> 
> I'm running the nVidia CUDA GPU application for s...@h using the anonymous
> platform. There is no CPU application specified for s...@h.
> 
> No other projects utilise the GPU.
> 
> My intention is that to maximise utilisation of the hardware, s...@h is to
> be run only on the GPU, whilst in parallel the other non-GPU projects
> are run on the CPU.
> 
> 
> On first trying the s...@h GPU application, all ran fine and continuously.
> 
> Now, after a few days, s...@h is running only periodically despite s...@h WUs
> queued up waiting to run.
> 
> The CPU ustilisation is less than 1% for the s...@h GPU application.
> 
> 
> Is this a case of s...@h debt freezing it out? Even though that makes the
> GPU idle?
> 
> 
> Regards,
> Martin
> 

_______________________________________________
boinc_dev mailing list
[email protected]
http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev
To unsubscribe, visit the above URL and
(near bottom of page) enter your email address.

Reply via email to