CUDA Superhero Challenge on Topcoder

http://www.topcoder.com/tc?module=Static&a…=NVIDIAOverview

aw man I bet I can’t compete :(

But I hear your next door neighbor, Tom Mirray, may be entering…

Where can I find the results for the CUDA Superhero Challenge?

I would like to see both the entries, and the winners.

I wonder if any of the programs will be posted for download? I would love to run a few, and still looking for an app to use both sides of my 295, and 280 at the same time to solve a problem. :thumbup:

The results are here http://www.topcoder.com/longcontest/stats/…ew&rd=13957 .

A clean-up version of my winning solution can be downloaded from http://www.povworld.org/michariser/binary/CCL.cu

Thanks for the post, and congratulations… :thumbup:

Could you post what the problem was…
Looks like you have to be a Top Coder member to find out.

Feel free to post anything as to why your solution worked out well for you.
I would love to hear anything you might have to say. I am a CUDA fan, and GPU acceleration in general. :)

I wish they would spotlight the entries in a video.

I feel like it’s a secret contest… :whistling:

This was not a secret contest… Almost many people knew this long time back…

The contest was to find the “Connected Component Length” - An image procesing algorithm. An image is given with lot of colours. One needs to identify the un-broken segments (of same color) and identify it as 1 component (named with the lowest x,y position of the component).

It is not that much parallel friendly… As you can see from bloodhunt’s code - one need to address the problem with multiple passes(different kernel launches)…

Thanks for the info…

I just found the video from the previous years…

http://www.topcoder.com/tc?module=Static&a…rom%20the%20TCO

Yes, you need to be registered at topcoder to read the full problem statement. It was about calculating CCL (http://en.wikipedia.org/wiki/Connected_Component_Labeling) of large images (up to 300’000’000 pixels). The second posting in this http://forums.topcoder.com/?module=Thread&…869&start=0 thread describes my solution. There is still some room for improvement in my solution, e.g. reading the image as texture instead from global memory.

Thanks for the link.

For future CUDA Superhero Challenges, I would rather not see a problem assigned.

I think the creative programs that might be submitted, will be more exciting for us nVidia fans to look at. They would also make for a better advertising video too.

This CUDA app is my current favorite: The CUDA Factorial Benchmark
http://www.xtremesystems.org/forums/showthread.php?t=233280
(It is also ready to be downloaded and run as is.) We need more apps like that. I downloaded your .cu file, but have no idea how to run it.

The CUDA Factorial Benchmark lets us see how much faster our GPU’s are, over our CPU. That is exactly the kind of app we need to help educate the non-believers in the GPU Revolution.

I want the version of the CUDA Factorial Benchmark that uses all 3 of my GPU’s, instead of just 1/2 of my 295. I have been told they are working on it. ;)

If anybody knows of a CUDA program that I can download, that will use both sides of my 295, and my 280 dedicated PhysX processor all at once to solve a problem, please post up a link!
I want that program real bad, whatever it is. :)

I had the incorrect idea that new apps were going to be submitted by CUDA programmers, that would be up for download, and that we could also run. That was bad on my part.
I assumed that was what the CUDA Superhero Challenge was all about.