Re: [AMBER] Need a higher value for SKINNB

From: Hailin Huang <hailin.huang.my.liu.edu>
Date: Wed, 19 Jun 2013 13:43:47 -0400

Hi Scott,

The "cudaMemcpy GpuBuffer" error came from the GTX 780 that Jonathan used
to post the test results. I switched to a 690 to run the same calculations
and everything works fine. What is this error about?

Thanks,
Richard


On Wed, Jun 19, 2013 at 1:16 PM, Scott Le Grand <varelse2005.gmail.com>wrote:

> I'll update the error condition to be a little more descriptive.
>
>
>
> On Wed, Jun 19, 2013 at 5:33 AM, Hailin Huang <hailin.huang.my.liu.edu
> >wrote:
>
> > Dear Jason,
> >
> > The problem was solved after I restarted the calculations, thank you!
> >
> > To one of my proteins (a mutant of the ones that run fine after restart),
> > the same error showed up a second time in my restart. So I restarted it
> > again, but this time I got this error "cudaMemcpy GpuBuffer::Download
> > failed unspecified launch failure". Still, this message showed up only in
> > the log file, the output just ended at a random step. Is recompiling
> Amber
> > 12 with the latest bug fixes the only way to solve this problem?
> >
> > Thanks in advance.
> >
> > Richard
> > _______________________________________________
> > 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 Jun 19 2013 - 11:00:04 PDT
Custom Search