LISSOM (http://homepages.inf.ed.ac.uk/jbednar/rflissom_small.html) is a model of human neocortex (mainly modeled on visual cortex) at a neural column level. The model was developed by Bednar, Choe, Miikkulainen, and Sirosh, at the University of Texas. I made different implementations of the model (with the help of Prof. Bednar), with the aim of porting it to GPU hardware using CUDA by nVidia. Now the library can run up to 9x faster than on modern CPUs. The project also involves usage of multiGPU systems and GPU clusters, to overcome the two big problems we face:

-lack of memory to store the network;

-computing power to simulate the model.

Presentation of the project

Research Abstract of the project