Why is GHC so much worse than JHC when computing the Ackermann function?

[Mikhail Glushenkov]

Hi,

On Sat, Apr 20, 2013 at 12:08 PM, Christopher Done <chrisdone at gmail.com> wrote:
> JHC compiles to C and last time I tried this it looked very much like the
> original C version:
> http://www.reddit.com/r/haskell/comments/1bcru7/damn_lies_and_haskell_performance/c9689a3
>
> This is the same thing.

Yes, but with the Fibonacci example GHC wasn't consuming all available
memory and running 10 times slower. Surely it must be possible to make
the GHC version at least 2x as slow as Ocaml/GHC?

-- 
()  ascii ribbon campaign - against html e-mail
/\  www.asciiribbon.org   - against proprietary attachments