[MITgcm-support] Single reduction CG solver for pressure

Martin Losch Martin.Losch at awi.de
Wed Jun 17 02:28:06 EDT 2009


Hi Christopher,

that would be cool, if you could speed up cg2d (and maybe even cg3d?).
As far as I know (but what do I know), your D'Azevedo scheme has not  
been tried. There's a cg2d_bench in the MITgcm_contrib area, maybe  
that's interesting for you?

Martin

On Jun 17, 2009, at 12:26 AM, Christopher L. Wolfe wrote:

>
> Hi modelers,
>
> I was just wondering if anyone has tried to implement D'Azevedo et  
> al.'s single reduction re-arrangement of the CG solver in cg2d. The  
> GC solver currently does three reductions per step---one of which  
> just checks the residual---and I find the model spends a lot of time  
> waiting for the reductions to complete (especially on the Cray XTs).  
> I think I may benefit from a rearrangement of the CG solver to  
> reduce the number of global reductions, but I don't want to reinvent  
> the wheel if someone else has already done it.
>
> Cheers,
> Christopher
>
> -----------------------------------------------------------
> Dr. Christopher L. Wolfe                   858-534-4560
> Physical Oceanography Research Division    OAR 357
> Scripps Institution of Oceanography, UCSD  clwolfe at ucsd.edu
> -----------------------------------------------------------
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support




More information about the MITgcm-support mailing list