Re: [AMBER-Developers] nofftw3 and MKL

From: Ross Walker <>
Date: Tue, 14 Oct 2014 00:34:44 -0700

>On Mon, Oct 13, 2014, T Luchko (Lists) wrote:
>> -nofftw3 is there because fftw3 requires a minimum of GNU 4.3 and Intel
>> 9. It used to be that the minimum compiler version we supported was GNU
>> 4.1.2. Iąd be happy to get rid of this restriction.
>I think we should assume compilers able to handle fftw3. I also note that
>pmemd uses -DPUBFFT even though fftw3 has been configured and is
>say with gnu compilers chosen. Does anyone know why this is?

I think this is mostly historical. If memory serves me correctly pmemd had
the option between PUBFFT and fftw v2. Although PUBFFT was generally
preferred since Bob's modifications to it made it actually faster than
FFTW for AMBER type workloads. I don't think anyone has taken the time to
update the interface to FFTW v3 which is why it sticks with PUBFFT - I
think... Perri put in MKL's FFT for the CPU based hybrid precision model
but I don't think that was added for double precision in the trunk - Perri?

If someone would like to volunteer to update PMEMD to support FFTW 3 and
test it (including performance) it would be much appreciated.

All the best

|\oss Walker

| Associate Research Professor |
| San Diego Supercomputer Center |
| Adjunct Associate Professor |
| Dept. of Chemistry and Biochemistry |
| University of California San Diego |
| NVIDIA Fellow |
| | |
| Tel: +1 858 822 0854 | EMail:- |

Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.

AMBER-Developers mailing list
Received on Tue Oct 14 2014 - 01:00:02 PDT
Custom Search