Skip site navigation (1) Skip section navigation (2)

Re: GPU Accelerated Sorting

From: Robert Haas <robertmhaas(at)gmail(dot)com>
To: Eliot Gable <egable(at)gmail(dot)com>
Cc: pgsql-performance(at)postgresql(dot)org
Subject: Re: GPU Accelerated Sorting
Date: 2010-09-21 19:27:59
Message-ID: AANLkTi=fxqadyQX07sXJjsXNxUGcRKYtVTer_PHCN9Bn@mail.gmail.com (view raw or flat)
Thread:
Lists: pgsql-performance
On Mon, Aug 30, 2010 at 9:46 AM, Eliot Gable <egable(at)gmail(dot)com> wrote:
> Not sure if anyone else saw this, but it struck me as an interesting
> idea if it could be added to PostgreSQL. GPU accelerated database
> operations could be very... interesting. Of course, this could be
> difficult to do in a way that usefully increases performance of
> PostgreSQL, but I'll leave that up to you guys to figure out.
>
> http://code.google.com/p/back40computing/wiki/RadixSorting

It would be hard to use this because, in addition to the fact that
this is specific to a very particular type of hardware, it only works
if you're trying to do a very particular type of sort.  For example,
it wouldn't handle multi-byte characters properly.  And it wouldn't
handle integers properly either - you'd end up sorting negatives after
positives.  You could possibly still find applications for it but
they'd be quite narrow, I think.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise Postgres Company

In response to

pgsql-performance by date

Next:From: OgdenDate: 2010-09-21 19:34:42
Subject: Re: Query much faster with enable_seqscan=0
Previous:From: Greg SmithDate: 2010-09-21 19:16:57
Subject: Re: Query much faster with enable_seqscan=0

Privacy Policy | About PostgreSQL
Copyright © 1996-2014 The PostgreSQL Global Development Group