On Sat, Dec 06, 2008 at 07:47:10PM +0000, Tony Travis wrote:
> I'm interested in the PS3 as a cheap way of getting my hands on a Cell 
> processor (PowerPC + 8SPE's). Ubuntu runs on it, and I'm interested in 
> finding out how good it is for bioinformatics. I nearly bought one at 
> half the price when I was in Canada, but it would not work in the UK.

There are a couple of projects using clusters of PS3's for various
things (I suspect you'll have seen most of these already):
http://gravity.phy.umassd.edu/ps3.html
http://moss.csc.ncsu.edu/~mueller/cluster/ps3/
http://www.scei.co.jp/folding/en/

I've spoken to a few people who've played with the cell for HPC and
have gotten very mixed opinions about it. I think it's possible to get
really good performance out of it if you have a problem that fits its
architecture (works on a few streams of data without too much
branching or random data access) but the code needs tweaking (or
completely rewriting) to get that performance and it requires thinking
a bit differently compare to current processors.

  Robert

________________________________________________________________________
Robert McWilliam            [EMAIL PROTECTED]             www.ormiret.com

Stress is when you wake up screaming and then realize you haven't
fallen asleep yet.

-- 
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-uk
https://wiki.ubuntu.com/UKTeam/

Reply via email to