24054Re: [PrimeNumbers] Re: ...factorizer to break RSA...
- Feb 15, 2012--- On Wed, 2/15/12, djbroadhurst <d.broadhurst@...> wrote:
> However, the GCD idea seems to work OK:Using an algorithm from (other) DJB. I said that might be the case, didn't I?
Aparently I'd only need to do ~400Mdigit GCDs rather than giga-digit ones, and less than half the RAM. That drops the estimation to about a day. Which is in remarkable agreement with that team's results. Most remarkable as it's a distinct algorithm from DJB's. It might be worth actually implementing, to see how the two compare to each other. To have apparently the same Big-Oh and similar constant to a DJB algorithm just from a back-of-a-fag-packet algorithm doesn't happen every day.
- << Previous post in topic Next post in topic >>