GTW’14 conference organized by Nanyang Technological University (Singapore) and NVIDIA has awarded a second runner up to Computational Neuroscience lab‘s work on GPGPUs on July 10, 2014 .
GPU Technology Workshop South East Asia (GTW-SEA) was the must-attend computing technology event focussing on opportunities to learn how to harness the latest GPU technology, along with face-to-face interaction with industry luminaries and NVIDIA experts.
The study was authored by Manjusha Nair, Bipin Nair and Shyam Diwakar and was titled “Realistic Large Scale Neuron Modeling and Simulations with GPU Acceleration”. This is the only Indian team in the award list.
The study involved mathematically modeling of cerebellar neurons and simulated execution of a large network via programming model imposed by General Purpose Graphics Processing Units such as those in CUDA architecture. Such models have a lot of relevance in mathematics and neuroscience. It helps make new hypothesis for neurobiologists and to test pharmacological protocols i.e. studying how drugs work. The goal of this model was also to understand how sensory signals are processed by the cerebellum. The model imitated a rat brain circuit since cerebellum is a brain structure that is fairly preserved in many species like rats and primates. The implementation also showed how information metrics like mutual information can be used to understand large entropic biological neural circuits.
The work was computed on Amrita School of Biotechnology‘s MANAS GPGPU supercomputer.
Another extended part of a similar work was also presented at the Eleventh International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics (CIBB), organized and hosted by University of Cambridge, UK, June 26-28, 2014. The paper was titled “Large-Scale Simulations of Cerebellar Microcircuit Relays using Spiking Neuron on GPUs”. The paper was part of the oral presentation session at CIBB 2014.
Manjusha Nair is also a faculty member at Amrita School of Engineering and is a part-time Ph.D. student at the Computational Neuroscience Lab at Amrita School of Biotechnology. Amrita School of Biotechnology was recognized as a NVIDIA Cuda Teaching Center from 2012-14.
