What is it about?
In this paper we tested the efficiency of our GPU based simulation framework for brain simulations in terms of speed and energy efficiency. While it is always difficult to make fair comparisons in this area, we did find that our framework did match or outperform recently published results obtained on a super-computer and on a neuromorphic system.
Featured Image
Why is it important?
Brain simulations are becoming increasingly important for understanding how the brain works and how it malfunctions in disease and disorder. However, current computers are still too slow and too small to allow detailed simulations of models approximating the size of the human brain. Scientists across the world are developing larger traditional and more unconventional computers to improve computational brain modelling. Our own framework for brain simulations uses GPUs which have only recently been established as scientific computing accelerators. This paper demonstrates that they are well-suited for brain simulations.
Read the Original
This page is a summary of: GPUs Outperform Current HPC and Neuromorphic Solutions in Terms of Speed and Energy When Simulating a Highly-Connected Cortical Model, Frontiers in Neuroscience, December 2018, Frontiers,
DOI: 10.3389/fnins.2018.00941.
You can read the full text:
Contributors
The following have contributed to this page