On Tuesday 18 December 2007 10:03:25 Dimitri Fontaine wrote:
> Hi,
>
> Le mardi 18 décembre 2007, Ron Mayer a écrit :
> > Has anyone looked into sorting algorithms that could use
> > more than one CPU or core at a time?
>
> [...]
>
> > PS: Yeah, I know multi-threading is a hot-button on these
> > lists; but sorting seems a relatively isolated of the code
> > and I'd wonder if it'd be isolate-able enough that multiple
> > CPUs could be used there.
>
> And before that objection to multi-threading implementation and portability
> concerns arise, what about using a coroutine BSD-licenced portable
> implementation such as Protothreads to have backend code use several CPU at
> a time?
> http://www.sics.se/~adam/pt/
>
> With such a tool, would it be possible to think about producer/consumer
> parallel executions for sorting, aggregates nodes or other parts of the
> executor?
>
> Hope this helps, regards,
And remember; Users don't care about portability-issues, they care about
performance. If multi-threading is a way to speed up sorting considerably, it
should, IMHO, be considered seriously.
--
Andreas Joseph Krogh
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
