Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!utgpu!water!watnot!watmath!clyde!cbatt!ihnp4!ptsfa!lll-lcc!seismo!husc6!bacchus!speter From: speter@bacchus.UUCP Newsgroups: comp.sys.ibm.pc Subject: MS-Fortran ver. 4.0 Message-ID: <274@bacchus.MIT.EDU> Date: Sun, 1-Mar-87 08:42:11 EST Article-I.D.: bacchus.274 Posted: Sun Mar 1 08:42:11 1987 Date-Received: Mon, 2-Mar-87 06:59:01 EST References: <1512@vax2.tc.fluke.COM> <16800056@clio> Sender: daemon@bacchus.MIT.EDU Reply-To: speter@athena.mit.edu (Peter Osgood) Organization: MIT Project Athena Lines: 67 Keywords: Info only... Group: comp.sys.ibm-pc I recently beta tested Microsoft Fortran version 4.0. To those of you who are serious fortran programmers this version is well worth a look see over the previous 3.31 version. It is on sale now and retails for about $450, less if you are a student at many colleges. They have managed to reduce their code size by 1/3 while increasing excution performance by almost 50%. I ran these tests on an IBM-PC XT w/640k, co-processor (8087), and hercules graphics. A simple 15 line code using floating point and a 30k repetition of a loop with an astro-physic algorithm included was used to bench mark against old 3.31 MS-Fortran and Ryan-MacFarland ProFortran. In beat both hands down. The answer, by the way, came out dead on every time to the minus 8 decimal. ---peter osgood--- Newsgroups: comp.sys.ibm.pc Subject: Summary: Expires: Sender: Reply-To: speter@athena.mit.edu (Peter Osgood) Followup-To: Distribution: Organization: MIT Project Athena Keywords: Group: comp.sys.ibm.p Subject: MS-Fortran ver. 4.0 Summary: Expires: References: <1512@vax2.tc.fluke.COM> <16800056@clio> Sender: Reply-To: speter@athena.mit.edu (Peter Osgood) Followup-To: Distribution: Organization: MIT Project Athena Keywords: Info only... I recently beta tested Microsoft Fortran version 4.0. To those of you who are serious fortran programmers this version is well worth a look see over the previous 3.31 version. It is on sale now and retails for about $450, less if you are a student at many colleges. They have managed to reduce their code size by 1/3 while increasing excution performance by almost 50%. I ran these tests on an IBM-PC XT w/640k, co-processor (8087), and hercules graphics. A simple 15 line code using floating point and a 30k repetition of a loop with an astro-physic algorithm included was used to bench mark against old 3.31 MS-Fortran and Ryan-MacFarland ProFortran. In beat both hands down. The answer, by the way, came out dead on every time to the minus 8 decimal. ---peter osgood--- Newsgroups: comp.sys.ibm.pc Subject: Summary: Expires: Sender: Reply-To: speter@athena.mit.edu (Peter Osgood) Followup-To: Distribution: Organization: MIT Project Athena Keywords: