Japan’s K computer, 4th largest supercomputer worldwide has managed to accurately map part of human brain activity in record time – 40 minutes. The machine simulated 1% of neuronal network using 705,024 processor cores and 1400 terabytes of RAM.
The project was designed to test the limits of the supercomputer in simulating human brain activity. The results from this exercise will be used to improve the Neural Simulation Technology (NEST), an open source platform used in the project. Researchers from Japanese RIKEN and the German Forschungszentrum Jülich centre joined in testing NEST.
“If petascale computers like the K computer are capable of representing one per cent of the network of a human brain today, then we know that simulating the whole brain at the level of the individual nerve cell and its synapses will be possible with exascale computers – hopefully available within the next decade,” – Markus Diesmann, part of the team conducting the simulation.
Despite replicating a network of 1.73 billion nerve cells and 10.4 trillion synapses, this simulation project represents a small contribution in the move to exascale computing. Power consumption is one of the challenges that lies in the way of exascale computing with experts noting that such machines can’t be created before the year 2020.