[comp.lang.fortran] Fortran to C Translator Reviewed

drd@siia.mv.com (David Dick) (06/30/90)

In <KHB.90Jun26114740@chiba.Eng.Sun.COM> khb@chiba.Eng.Sun.COM (Keith Bierman - SPD Advanced Languages) writes:


>>In Digital Review (June 18, 1990) there is a review of Promula.Fortran, a
>>Fortran to C source code translator on pages 29 - 31.

>An important thing to note is that the reviewer obserseved that the
>runtimes of the translated code was about .6 the original fortran times.
                                            ^
					    |
Isn't this better?!-------------------------+
What is the real story?

>The reviewer tracked down some of the most painful slowdowns and
>blames them on the VAX C compiler ... not on C features. 

David Dick
Software Innovations, Inc. [the Software Moving Company(sm)]

doug@jhunix.HCF.JHU.EDU (Douglas W O'neal) (07/02/90)

In article <1990Jun29.225129.5160@siia.mv.com> drd@siia.mv.com (David Dick) writes:
->In <KHB.90Jun26114740@chiba.Eng.Sun.COM> khb@chiba.Eng.Sun.COM (Keith Bierman - SPD Advanced Languages) writes:
->>>In Digital Review (June 18, 1990) there is a review of Promula.Fortran, a
->>>Fortran to C source code translator on pages 29 - 31.
->>An important thing to note is that the reviewer obserseved that the
->>runtimes of the translated code was about .6 the original fortran times.
->                                            ^
->Isn't this better?!-------------------------+
->What is the real story?

Real story: The benchmark suite that DR uses gave a Vaxstation II/GPX rating
of 0.61 MVUPs (MicroVAX II units of processing).  Running the original code
on the same machine would give a rating of 1.0 MVUPs.  Quoting from page 31:
    "We found that the translated C code ran, at best, neck-and-neck with 
     the Fortran originals, but on average was 0.61 times as fast.  The
     worst-performing kernel was GAMSIM, which ran 0.27 times as fast as
     its Fortran counterpart."
-- 
Doug O'Neal, Distributed Systems Programmer, Johns Hopkins University
doug@jhuvms.bitnet, doug@jhuvms.hcf.jhu.edu, mimsy!aplcen!jhunix!doug 
Like many of the features of UNIX, UUCP appears theoretically 
unworkable... - DEC Professional, April 1990