Sisal? 
Author Message
 Sisal?

Where might I find discussions/critiques/rebuttals regarding Sisal?
Is this dataflow/single-assignment language "functional", and how does it
manage to beat fortran?  I have already looked at sisal.llnl.gov.

--




Sun, 02 Mar 1997 22:46:26 GMT  
 Sisal?
    >> Where might I find discussions/critiques/rebuttals regarding
    >> Sisal?  Is this dataflow/single-assignment language
    >> "functional", and how does it manage to beat Fortran?

It only beats FORTRAN on multi-processor parallel machines. It is first
order, strict, and mono-morphic, however still a functional language. Most
functional languages are easier to put into a parallel form than FORTRAN.
Also as Preston said, the global optimization is simpler for FLs. Sisal
even beats "c" on my HP710 workstation. I tried the "pseudo-knot"
benchmark (numerical) and Sisal was 30-40% faster even though I was using
the same "gcc" compiler! The other parallel (already) functional language
that was in the same benchmark suite was Concurrent-Clean. This one works
with native code so it is available only on a few platforms (linux,
macintosh, suns). It is very fast lazy, Milner typed, modern I/O. Sisal is
going to be upgraded to version 2.0 (soon?). This new version will be
polymorphic, higher order functions,.. There is a paper describing the new
version at /pub/sisal::sisal.llnl.gov as well as examples, current
implementations.  There is even a paper "Is FORTRAN dead?" ! The examples
are large programs (one for Weather simulation) originally written in
Fortran.

Osman

--




Fri, 07 Mar 1997 20:27:46 GMT  
 Sisal?
: Where might I find discussions/critiques/rebuttals regarding Sisal?
: Is this dataflow/single-assignment language "functional", and how does it
: manage to beat Fortran?  I have already looked at sisal.llnl.gov.

On a related note ....  Has anybody tried to build a FORTRAN to Sisal
(source-to-source) compiler.  Would this be considered useful in a
parallel programmng environment?  Any thoughts?

- edward
-------------------------------------------------------------------
Edward Walker
National Supercomputing Research Centre
81, Science Park Drive
#04-03, The Chadwick
Singapore 0511
SINGAPORE




Sat, 08 Mar 1997 17:27:42 GMT  
 Sisal?

Quote:
>> Where might I find discussions/critiques/rebuttals regarding
>> Sisal?  Is this dataflow/single-assignment language
>> "functional", and how does it manage to beat Fortran?
>It only beats FORTRAN on multi-processor parallel machines. It is first

I hope you'll pardon me for taking some of the wind out of your sails,
but about 5 minutes ago, I completed a benchmark of a simple convolution
code on the SUN-4, comparing SISAL, Fortran, and my APL compiler.

For a fairly hefty convolution (250x4000), I observed times as follows:

F77 -O3                           0.730u   0.180s
SISAL -O -nobounds -cc="-O4"      0.610u   0.090s

I'm not ready to announce APL results yet.

Note that this is a vanilla, single processor machine, so your claims
about "only... multiprocessor.,." ain't strictly true.

Basis for comparison: I tried to get the most optimized version of both
fortran and SISAL. I am not that familiar with the SUN-4, and it may be
that there are better-performing compiler options. If so, please let me
know.

One note of interest: I had a fancy, hand-unrolled (10) code, originally
destined for a CRAY X-MP, and it ran SLOWER on the SUN-4 than did the
naive loop. May have been unrolled too far? I don't know.

Bob
--




Mon, 10 Mar 1997 06:43:07 GMT  
 Sisal?

Quote:
>[re Sisal vs. Fortran performance]
>>It only beats FORTRAN on multi-processor parallel machines. It is first

>I hope you'll pardon me for taking some of the wind out of your sails,
>but about 5 minutes ago, I completed a benchmark of a simple convolution
>code on the SUN-4, comparing SISAL, Fortran, and my APL compiler.

>For a fairly hefty convolution (250x4000), I observed times as follows:

>F77 -O3                           0.730u   0.180s
>SISAL -O -nobounds -cc="-O4"      0.610u   0.090s

Let me take some of the wind out of my own sails. I received two
messages from folks who suggested:
   f77 -fast -dalign -O4

This produced code which ran neck and neck with SISAL.

Bob
--




Tue, 11 Mar 1997 12:15:31 GMT  
 
 [ 5 post ] 

 Relevant Pages 

1. SISAL Developers' Workshop at PACT'97

2. looking for sisal 2 manual

3. Sisal lives

4. Sisal still available?

5. what happened to Sisal?

6. Who know If1 for Sisal?

7. What has happened to Sisal?

8. SISAL Developer's Workshop Notes

9. PACT-97 with Sisal Workshop

10. SISAL Developers' Workshop at PACT'97

11. Calling SISAL programs from C

12. SISAL as Intermediate Language for parallel compilers

 

 
Powered by phpBB® Forum Software