GPUs for Genetic and Evolutionary Computation contest

Accelerated Computing CUDA CUDA Programming and Performance
1

At GECCO 09 today, the GPU computing contest received a surprisingly robust set of 10 quality entries. These all included source and executable code and a two page writeup. One is my own entry in fact.
The entries are now available on the contest page.
The goal is to show off applying new evolutionary algorithms to the GPU, with points based on utility, generality, efficiency, and originality.
The judging results are still unknown but you can now read through the entries via the link above.

2

Sweet, all but one uses CUDA :)

3

Was this contest ever announced here?

Must have missed it, if it was.

Christian

4

Make sure the NVidia marketing guys know that. I always laugh at online GPU reviews or discussions where to be “fair” the authors present AMD’s CTM as being a roughly equivalent toolset as CUDA.

5

It was on the front page of GPGPU.org for a while. That’s where I saw it, and spent a rather fun weekend hacking together an entry from some old tests I had done once.

6

Hi, I am one of the co-organisers of that competition.

Today, we announced the top three papers based on the referee’s scoring:

  • Nicholas A. Sinnott-Armstrong, Casey S. Greene and Jason H. Moore - Using Evolutionary Computing on Consumer Graphics Hardware for Epistasis Analysis in Human Genetics

  • Steve Worley - Optimization of Primality Testing Methods by GPU Evolutionary Search

  • Luca Mussi and Stefano Cagnoni - Particle Swarm Optimization within the CUDA Architecture

The entrants were able to give a short talk, and a ballot was given out to pick the final winner. This will be announced on Sunday. (I will update the webpage on Sunday too).

Also at the conference we did a special session and a tutorial. Both were well attended and well received.

I hope we can hold the content again next year. The entries were of extremely high standard and reviewing was a pleasure. Thank you to all who submitted.

7

I should also thank NVidia for sponsoring the contest and helping with reviews!

8

Congrats to all the winners! Glad to hear it went well, Simon.

9

The winning paper was announced today:
Nicholas A. Sinnott-Armstrong, Casey S. Greene and Jason H. Moore - Using Evolutionary Computing on Consumer Graphics Hardware for Epistasis Analysis in Human Genetics

Congratulations to all involved!