Is there a platform optimized version of gmp for INTEL?

[Brian Gladman]

Michael wrote:
> Does anybody have a statistic about how much performance is gained by using
> Intel Compiler and Intel tools, than MSVS 2005?

In my experience with the Intel and Microsoft compilers on 32 and 64 bit
x86 systems, it is impossible to predict which will provide faster code
without simply trying them both on a particular application.

In the case of GMP, as others have said, unless you are using the
generic C code without assembler, I would not expect the compiler to
make a big difference for most applications.

> Also if I pay big bucks and hire some experts to rewrite some portions of
> the GMP and MPFR (only a small portion that is related to my project, to
> make the whole thing manageable) in a Intel platform optimized Assembly
> language. Will that improve speed a lot?

Before you do this it would be useful to know where the bottlenecks in
your application are and the extent to which GMP performance plays a
part in them.

GMP contains assembler for routines that are likely to be heavily used
but the extent of assembler support varies widely between different
platforms. If your application makes extensive use of a GMP routine for
which assembler support is not present on your platform, then the
provision of this routine in assembler might give you a significant
performance gain.

Do you have evidence indicating which GMP routines are critical for your
application and the extent to which these are making use of assembler
support on your platform?

   Brian Gladman