GMP on the Cell processor

[Linas Vepstas]

On Tue, Apr 17, 2007 at 09:18:22PM +0100, Paul Leyland wrote:
> 
> I've been giving it some thought recently, but the pressure of Real Life
> (TM) and, especially, Real Work has been such that the thought hasn't
> got very far yet.

I need to say this again: do not be mislead my my email address, as 
I'm only looking at this as a possible skunk-works in parallel to
Real Work (TM).

> My interest is primarily in integer factorization, several algorithms

FWIW, there are several efforts looking at FFT on Cell, with at least
one of them being free as in freedom, ad mostly all done.

> for which are trivially parallelizable and computationally demanding (as

This fact has not been lost on certain govt agencies. A back of the
envelope calculation shows that, for about $250K, you can fill the 
garage of a suburban home with PS3's, power them without requiring
an electrical utility upgrade from the city, and break 64-bit DES
keys in about a week or two. All this without arousing suspicion 
by actually renting a space in an industrial park ... 

--linas