Re: A very fast Fermat factoring algorithm

From: Tom St Denis (tomstdenis_at_gmail.com)
Date: 03/30/05 

Date: 30 Mar 2005 09:36:25 -0800

quantumgecko wrote:
> I'd rather not share the technique just yet. I don't know for sure
if
> it has any value.

What do you mean value... First off 10^-9 compared to 10^300 [the size
of an RSA-1024 bit key] is nothing.

Roughly speaking if the Fermat method takes O(n^1/4) steps [2^256 for
1024-bit number] then your method reduces the work to 2^226 steps.
Which is roughly
1,393,796,574,908,163,946,345,982,392,040,522,594,123,776 TIMES harder
than GNFS.

So your method or idea may have academic merit but it won't be breaking
RSA keys before I go to bed tonight.

Tom