Yes, it would be interesting to see some of the benchmarks from the parallel nofib used, http://www.macs.hw.ac.uk/~dsg/gph/nofib/ simonpj:
Friends: an opportunity spotted by Ulf.
Simon
-----Original Message----- From: ulf.wiger@gmail.com [mailto:ulf.wiger@gmail.com] On Behalf Of Ulf Wiger Sent: 12 August 2008 16:33 Subject: shootout on quad-core
The famous language shootout now has a quad-core architecture:
http://shootout.alioth.debian.org/u64q/benchmark.php?test=all&lang=all
So far, the maintainers have mainly focused on trying to get the build environment to work, and no real agreement exists on how to potentially expand the benchmark suite with multicore problems, or perhaps how to handle alternative entries on existing benchmarks.
I think they welcome suggestions. I imagine that several different categories exist:
- problems that naturally lend themselves to message-passing concurrency - obviously data parallel algorithms - problems that could be parallelized using a smart compiler
BR, Ulf W
_______________________________________________ Haskell mailing list Haskell@haskell.org http://www.haskell.org/mailman/listinfo/haskell