Roman Leshchinskiy schrieb:

I'm pleased to announce NoSlow, a nascent benchmark suite for various array (and list) libraries with particular emphasis on finding out how well they fuse things. At the moment, it knows about

* standard lists * primitive DPH arrays (dph-prim-seq) * uvector * vector * storablevector

It compiles and runs (using Brian's criterion) a fairly random collection of very small loop kernels for each of those and produces a lot of data which it then uses to generate ugly HTML tables. In the future, it will have more benchmarks, more complex benchmarks and much prettier tables.

You can get more information (including the ugly tables) from my blog

http://unlines.wordpress.com/2009/11/27/noslow

and NoSlow itself from Hackage

http://hackage.haskell.org/package/NoSlow

Any help would be highly appreciated as I don't have a lot of time to work on it.

Since storablevector is in no way optimized for fusion, I'd like to see how well it compares with Fusion-Streams that are fused into strict and lazy storablevectors by http://hackage.haskell.org/package/storablevector-streamfusion . Thanks for the benchmarking effort!