Fixed. I meant NPT - as in if we choose settings that use the most GPU memory how many atoms (approximately) will fit in a given amount of GPU memory space.
> On Nov 18, 2015, at 08:20, Jason Swails <jason.swails.gmail.com> wrote:
>
> On Wed, Nov 18, 2015 at 10:25 AM, Robert Molt <rwmolt07.gmail.com> wrote:
>
>> I am confused. The manual states, on p. 310, that ntb=1 is constant
>> volume. On p.307, the ntp command, when it has a value other than ntp=0,
>> makes constant pressure. In the performance listing, on
>> http://ambermd.org/gpus/ for explicit solvent PME, it says it uses
>> NTB=1, NTP=1. This would lead me to believe constant pressure, constant
>> volume (and ntt=3 makes constant temperature).
>>
>
> It's a typo. pVT is clearly not a real ensemble.
>
> Ross -- look around http://ambermd.org/gpus/#system_size_limits for what
> Robert is talking about. It's not immediately clear to me whether you
> meant NpT or NVT.
>
> That said, treat these as basic ball-park estimates and you'll be fine.
> NVT can squeeze a few more atoms into memory, but unless you're operating
> near that limit (i.e., with millions of atoms), you don't have to worry
> about this distinction.
>
> HTH,
> Jason
>
> --
> Jason M. Swails
> BioMaPS,
> Rutgers University
> Postdoctoral Researcher
> _______________________________________________
> 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 Thu Nov 19 2015 - 08:30:04 PST
