Re: PHD subject

From: Colin Percival (cperciva_at_sfu.ca)
Date: 10/12/05 

Date: Wed, 12 Oct 2005 17:57:32 +0000 (UTC)

Pubkeybreaker <Robert_silverman@raytheon.com> wrote:
> What would be most desirable is a *parallel* method that would
> allow one to partition the matrix, perform computations on each
> partition *uncoupled*, then paste the results together at the
> end.

I think the *most* desirable option would be a psychic who can look
at sparse matrices and write down the solution -- but yes, a parallel
algorithm would be very nice. :-)

> Note: I have looked for such a technique......

I think everybody has...

Colin Percival