Re: Number of Particles; Exponent 2?

From: Gabriel Altay <gabriel.altay_at_gmail.com>
Date: Thu, 21 Aug 2008 08:59:02 -0400

I would imagine that having a number that is a factor of 2 would only be
relevant for the TreePM option and would apply to the choice of PMgrid and
not Npar. Many people use a factor of two more grid cells than they do
particles which is probably (one of) the reason(s) why the factors of 2
appear so often in the particle loads. Even so, the FFTW libraries perform
reasonably well with any PMgrid that has small prime factors.

On Thu, Aug 21, 2008 at 6:46 AM, Sami Niemi <saniem_at_utu.fi> wrote:

>
> Hi All Gadget2 Users,
>
> Does anyone know if deviation from power of two in base number of particles
> have a meaningful effect on running time? From literature it is quite
> obvious that most groups are using a base number of particles that is
> exponent two (e.g. 512**3). However, when the number of particles increases
> above, lets say, 1024**3 it would be nice to use arbitrary number of
> particles for example 1500**3. Can anyone confirm that there is no any
> additional delay when applying 1000**3 compared to 1024**3 particles?
>
>
> Best Regards,
> Sami
> _______________________
> Sami-Matias Niemi
> Student Support Astronomer
> Nordic Optical Telescope
>
> sami_at_not.iac.es
> +34 662 535 441
> +34 922 425 424
>
> http://users.utu.fi/saniem/
> http://saniem.deviantart.com
> ----------------------------------------
>
>
>
>
> -----------------------------------------------------------
> If you wish to unsubscribe from this mailing, send mail to
> minimalist_at_MPA-Garching.MPG.de with a subject of: unsubscribe gadget-list
> A web-archive of this mailing list is available here:
> http://www.mpa-garching.mpg.de/gadget/gadget-list
>
Received on 2008-08-21 14:59:05

This archive was generated by hypermail 2.3.0 : 2023-01-10 10:01:30 CET