On May 7, 2014, at 5:46 AM, <ahmed.sajid.stfc.ac.uk> <ahmed.sajid.stfc.ac.uk> wrote:
> Hi Guys,
>
> Is anybody aware of this issue being resolved?
I haven't heard anything about NVidia fixing this, but there is a simple workaround described in recent post on the Amber mailing list. If you google "cudaDeviceEnablePeerAccess all CUDA-capable devices are busy or unavailable" it comes back with:
http://archive.ambermd.org/201403/0127.html
You can try the latest NVidia driver, but if that doesn't work then the above workaround is your best bet.
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed May 07 2014 - 05:00:04 PDT