This project has moved. For the latest updates, please go here.

just submitted 2 new implementations.

Jun 29, 2011 at 7:29 AM

would have been nice to try Strassen's algorithm, but both implementations have a respectable speed up w/o it

Jun 30, 2011 at 9:04 AM

Hey Stephen,

Looks like it might just be you and me :-)

Bummer about Strassen - I was hoping someone else would be inspired enough to do it so could plug into my solution!

Anyway, I guess we'll find out soon enough - although they said a winner wouldn't be announced until the end of July, so maybe I can just post my code up somewhere tomorrow (considering the comp should be closed), if you are keen to check it out. (I would be keen to see what you have done, if you are willing)


Jul 7, 2011 at 3:46 PM

yeah strassen's would be nice to implement to see what sort of speedup it gives, though it looks like most people implement it as part of another solution. When I was thinking about it, I ran some numbers in Excel and it doesn't start to yield a nice return (in terms of the number of computations) until n=300 or so.

I'd love to trade code w/ you. I forked the project onto bitbucket when I started (private repos) but I could fork the contest code onto codeplex and port my changes over if you're willing to do the same

Jul 23, 2011 at 1:03 AM

Argh! Either I don't have email linked to this account, or it's been sent to spam, as it was only now heading back here do I see your reply. (Actually just as I finish typing this email I see a checkbox at the bottom for return email...)

Sounds like an excellent idea.

Here or bitbucket are fine. My bitbucket account name is "stickman".