[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index]
[Thread Index]
- Subject: Re: [ANN] LuaJIT-2.0.0-beta1
- From: steve donovan <steve.j.donovan@...>
- Date: Sun, 1 Nov 2009 11:09:01 +0200
On Sun, Nov 1, 2009 at 12:59 AM, Mike Pall <mikelu-0910@mike.de> wrote:
> SciMark composite |
> small score | FFT SOR MC SPARSE LU
> ----------------------------+---------------------------------------
> GCC 4.3.2 906.1 | 739.1 909.0 190.4 1057.0 1635.1
...
> JVM 1.6 Client 579.6 | 424.8 895.8 122.8 595.5 859.0
> ----------------------------+---------------------------------------
> LuaJIT 2.0.0-beta1 580.4 | 427.4 1025.0 223.7 303.4 922.5
What is particularly impressive, from the point of view of someone who
has to do serious numerical work occaisionally, is that LJ2 is right
up there with the compiled languages for raw performance - and this is
just the beta! So, the generalization 'dynamic languages are slow' is
now officially outdated. Also, I would suspect that the memory use
for LJ2 versus Java would be much less. (So much for another common
generalization, that one has to pay for VM use in memory)
So, congratulations are in order, as well as the suggestion that the
phenomenon named 'Mike Pall' definitely needs more funding!
Anyone else who thinks it's time for an MPI binding for Lua?
steve d.