Re: [gnugo-devel] Parallelizing GNU Go

[bump]

> I am taking a parallel programing class at UT Austin, and I was
> considering parallelizing GNU Go for my class project.
> 
> I have a lot of thoughts on how to go about doing this, but first I wanted
> to see what people thought of the idea in general.
> How difficult would it be?
> Is there anyone who's already working on such a project?

We'd like to hear your ideas.

If anyone is working on this we are not aware of it. It is not
obvious how to start without substantially revising the
architecture of the program. Nevertheless this is likely
to be an important step.

An ambitious idea is to parallelize GNU Go in the context of
full board search, something that we currently don't do. If 
this is what you have in mind, there will be major revisions
to be made. Full board search will almost certainly have to
be done before one can write a truly strong program.

A less ambitious idea would be to run the owl reading for
different dragons in parallel. This would allow one to
increase the number of owl nodes and other improvements
which would currently be too slow.

Dan