NEC-LIST: Compiling NEC for a Large Number of Segments

Dear CEMists,

Thank you all for your replies. I am investigating John Blackburn's
suggestions re the g77 compiler, but at present favour the Intel approach
because of the dramatic speed improvement to NEC when using the LAPACK
routines from the Intel Math Kernel Library. This used to be a free
download and included libraries for both Compaq Visual Fortran (CVF) and
Intel Fortran. Now one has to buy it along with new Intel Visual Fortran
(as Compaq Visual Fortran is no more, with IVF being the official upgrade
route.)

However, the Intel web site is far from clear on exactly whether the current
versions of IVF and MKL can take full advantage of the Intel P4 64-bit
processor. I asked a question on the Intel Forum and Steve Lionel of Intel
replied as follows:

If you run Windows x64 and use the "Intel EM64T" 64-bit compiler, you can
have ALLOCATABLE arrays larger than 2GB. Even on Windows x64, static arrays
(declared as local variables or in COMMON) are limited to 2GB - that is a
Windows restriction.

For example, instead of this:

real bigarray(4000000000)

you would have this:

real, allocatable :: bigarray(:)
...
allocate (bigarray(4000000000))

The rest of the code would not need to change. If the big array is in
COMMON, you'll need to change to using modules and make the array a module
variable. Let us know if you need help with that.

I have asked a further question as to exactly what tools (and any add-ons)
one needs from Intel and MS. (IVF needs MS Visual Studio, but with certain
64-bit add-ons, I believe). I will keep you posted on the answer.

Regards,

Paul Carlier

-- 
The NEC-List mailing list
NEC-List_at_robomod.net
http://www.robomod.net/mailman/listinfo/nec-list