Re: Fortran 77 parser
- From: Janne Blomqvist <foo@xxxxxxxxxxx>
- Date: Mon, 7 Apr 2008 10:56:34 +0300 (EEST)
On 2008-04-06, Ron Shepard <ron-shepard@xxxxxxxxxxxxxxxxxx> wrote:
Even today, most compilers do not use the fast DGEMM() code for the
intrinsic MATMUL() function, even when the compiler vendors
distribute the BLAS library along with the compiler, but that is a
second-order tangent to this thread.
Following this tangent, gfortran 4.3 has a -fexternal-blas option that
causes it to use *GEMM for MATMUL (above a threshold whose default
value can be changed with yet another option). However, there is no
BLAS library distributed with the compiler, the user has to link in a
suitable one (it was thought to be better that way since gfortran
supports so many platforms; it was not considered feasible to
distribute platform-optimized BLAS libraries).
[SCILIB]
Currently Cray provides a SCILIB for their XT series, but it seems to
have little in common with the SCILIB described in this thread. IIRC
the modern SCILIB is just their packaging (with perhaps some
optimizations?) of SuperLU, SCALAPACK and BLACS.
--
Janne Blomqvist
.
- References:
- Re: Fortran 77 parser
- From: Jon Harrop
- Re: Fortran 77 parser
- From: glen herrmannsfeldt
- Re: Fortran 77 parser
- From: James Giles
- Re: Fortran 77 parser
- From: Richard Maine
- Re: Fortran 77 parser
- From: James Giles
- Re: Fortran 77 parser
- From: *** Hendrickson
- Re: Fortran 77 parser
- From: James Giles
- Re: Fortran 77 parser
- From: Walter Spector
- Re: Fortran 77 parser
- From: James Giles
- Re: Fortran 77 parser
- From: Walter Spector
- Re: Fortran 77 parser
- From: Ron Shepard
- Re: Fortran 77 parser
- From: James Giles
- Re: Fortran 77 parser
- From: Ron Shepard
- Re: Fortran 77 parser
- From: Walter Spector
- Re: Fortran 77 parser
- From: Ron Shepard
- Re: Fortran 77 parser
- From: Walter Spector
- Re: Fortran 77 parser
- From: Ron Shepard
- Re: Fortran 77 parser
- Prev by Date: Re: UPdate all Dependencies(CVF)
- Next by Date: Re: UPdate all Dependencies(CVF)
- Previous by thread: Re: Fortran 77 parser
- Next by thread: Re: Fortran 77 parser
- Index(es):