Re: Sisal?

wchang@genome.lbl.gov (William Chang) writes:
>Where might I find discussions/critiques/rebuttals regarding Sisal?
>Is this dataflow/single-assignment language "functional", and how does it
>manage to beat Fortran?


Unfortunately, I don't have any handy references for Sisal. However,
I can talk to the last question from memory.


In one of the Sisal papers, they discuss the language and an
implementation. They also offer some speed comparisons of programs
written in Sisal versus programs written in Fortran (I believe running
on a Cray). In some (all?) cases, the Sisal versions were faster.


They claimed the different versions of the programs used the same
algorithms and were coded naturally, so that much is probably fair.
However, the compilers were not equivalent, in that the Sisal compiler
examined the entire program, performing significant interprocedural
optimizations (and eventually emitting C to be compiled by the native
C compiler). The Fortran compiler, on the other hand, used the
traditional approach to separate compilation and was thus much more
restricted in the scope of its analysis and transformation. In
practice, we would also expect the Fortran compiler to be much more
usable, since it doesn't require recompilation of the entire program
after every edit.


Nevertheless, the results are impressive and have attracted attention
(and users!). While I don't have references to give you, there have
been several papers published. Hopefully someone else will have them
handy.


Preston Briggs
--

Return to the comp.compilers page.
Search the comp.compilers archives again.