Re: [AMBER] "cudaDeviceEnablePeerAccess all CUDA-capable devices are busy or unavailable"

From: Scott Le Grand <varelse2005.gmail.com>
Date: Wed, 7 May 2014 10:31:16 -0700

Are you running more than one AMBER process per GPU? If so, consider this
a feature. I have a "fix" for it, but it's going to tell you that you're
doing something you shouldn't be doing. All running more than one process
per GPU will achieve is slowing things down.


On Wed, May 7, 2014 at 8:41 AM, Ross Walker <ross.rosswalker.co.uk> wrote:

> Hi Ahmed,
>
> Can you provide more details here please? Are you referring to AMBER 14
> running across multiple GPUs or are you referring to AMBER 12?
>
> Without context it is difficult to understand what you mean by the 'issue'
> and being 'resolved'.
>
> All the best
> Ross
>
>
>
> On 5/7/14, 2:46 AM, "ahmed.sajid.stfc.ac.uk" <ahmed.sajid.stfc.ac.uk>
> wrote:
>
> >Hi Guys,
> >
> >Is anybody aware of this issue being resolved?
> >
> >Regards,
> >Ahmed.
> >
> >--
> >Scanned by iCritical.
> >
> >_______________________________________________
> >AMBER mailing list
> >AMBER.ambermd.org
> >http://lists.ambermd.org/mailman/listinfo/amber
>
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed May 07 2014 - 11:00:03 PDT
Custom Search