[erlang-questions] widefinder update

Hynek Vychodil <>
Mon Oct 29 11:35:05 CET 2007


Ruby code is slower on my old single core home desktop and sys time
are almost same.

http://pichis-blog.blogspot.com/2007/10/faster-than-ruby-but-scalable.html

--Hynek (Pichi) Vychodil

On 10/29/07, Thomas Lindgren <> wrote:
>
> --- Steve Vinoski <> wrote:
>
> > On 10/28/07, Anders Nygren <>
> > wrote:
> > >
> > > On 10/28/07, Hynek Vychodil
> > <> wrote:
> > > > Hi Anders,
> > > > I rewrote your code a little. I removed all
> > remaining binary bindings
> > > > and it is noticeable faster again. Try
> > wf_pichi3.erl.
> > > >
> > >
> > > Hynek
> > > that was great, Your change brings my wfinder1_1 +
> > wfbm4_ets1_1
> > > down to
> > > real    0m1.118s
> > > user    0m1.640s
> > > sys     0m0.368s
> > > on my 1.66 GHz dual core laptop.
> >
> >
> > And on the 8-core 2.33GHz Intel Xeon Linux box with
> > 2 GB RAM, this version
> > is extremely fast:
> >
> > real 0m0.567s
> > user 0m2.249s
> > sys 0m0.956s
>
> (I'll ignore the unexplained sys time below. That
> makes the discussion a bit preliminary; perhaps the
> derived results should be computed some other way.
> Apply grain of salt appropriately.)
>
> For those keeping track, the latest result is fully
> 2.7 times faster than the best previous version (which
> was block read), and 17.3 times faster than the
> initial version. The latest speedup is basically due
> to doing less work. However, note that user time fell
> by a somewhat greater ratio than real time, which
> might mean parallelization overheads are becoming
> visible.
>
> Also, the user time of 2.249 seconds is now close to
> the Ruby user time, which were 2.095s on the same
> hardware, while the Erlang parallelization speedup
> (user/real) on top of this is 3.95 out of 8. Comparing
> the real times of Erlang (0.567s) and Ruby (2.21s), we
> get about the same execution time speedup, 3.9. Not
> too shabby, huh?
>
> Is there anything more to be wrung out of this
> program? Well, apart from further tuning, one can note
> that Ruby had 0.1s sys time, while Erlang apparently
> needs a bit more, 0.5-1.0s. Why?
>
> Taking a wider view, it would also be very interesting
> to see how to apply these lessons to more general
> problems and/or libraries.
>
> Best,
> Thomas
>
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
> _______________________________________________
> erlang-questions mailing list
> 
> http://www.erlang.org/mailman/listinfo/erlang-questions
>



More information about the erlang-questions mailing list