This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: 128-bit precision
- From: "Jari & Tim Prince" <tprince at computer dot org>
- To: "Benjamin Herbert" <herbert at isis dot visi dot com>, <gcc at gcc dot gnu dot org>
- Date: Thu, 5 Sep 2002 18:45:34 -0700
- Subject: Re: 128-bit precision
- References: <20020905195110.GB7844@isis.visi.com>
----- Original Message -----
From: "Benjamin Herbert" <herbert@isis.visi.com>
To: <gcc@gcc.gnu.org>
Sent: Thursday, September 05, 2002 12:51 PM
Subject: 128-bit precision
>
> One of my researchers wants to test algorithms on a linux x86
> box that will eventually be run on a cray. The cray supports 128-bit
> precision arithmetic operations. He wants to use namely fortran, but
> also C and C++. Will gcc on a linux x86 machine support 128-bit
> precision arithmetic operations? If not does anybody know of a
> commercial product that can support what I am looking for.
>
gcc/g++ support the native x87 80-bit long double; g77 does not.
Several excellent commercial Fortran compilers support the 128-bit quad
precision, including Intel and Lahey. It will be quite slow. see
www.polyhedron.com