distributed & high performance programming trends

[D Herring]

Hi all,

Tonight I was telling some people about programming trends I was
seeing in the wider community. The Lisp community has a good
foundation for writing the necessary tools, but the work is being done
in other languages for various pragmatic reasons.

Here are a two of those languages, with papers motivating the problem
domain and outlining their approach to the solution.

Chapel:
http://chapel.cray.com/
http://psc.informatik.uni-jena.de/teach/prog-lang/HPCS/chapel-programmability.pdf

Halide:
http://halide-lang.org/
http://people.csail.mit.edu/jrk/halide12/halide12.pdf


Later,
Daniel

[Frank DG1SBG]

D Herring <dher...@at.tentpost.dot.com> writes:

> Hi all,
>
> Tonight I was telling some people about programming trends I was seeing in the
> wider community. The Lisp community has a good foundation for writing the
> necessary tools, but the work is being done in other languages for various
> pragmatic reasons.
>
> Here are a two of those languages, with papers motivating the problem domain
> and outlining their approach to the solution.
>
> Chapel:
> http://chapel.cray.com/
> http://psc.informatik.uni-jena.de/teach/prog-lang/HPCS/chapel-programmability.pdf

This looks interesting. I'd be inclined to make an attempt at
implementing Chapel as a set of Lisp macros, based on bordeaux-threads
and a few new data management utilities. Any co-workers here?

Thanks for sharing, Daniel!

Regards
Frank

[Marco Antoniotti]

Cool stuff. Unfortunately DARPA preferred Chapel over Fortress (and GLS wasted too much energy of LaTeX >:) ). Yet Fortress is worth a look.

MA