The HPE supercomputer will help Swiss researchers in the field of brain research



[ad_1]

Hewlett Packard Enterprise (HPE) created a "next generation" supercomputer for mammalian brain modeling and simulation.

The technology company has brought its experience in high performance computing (HPC) to the Blue Brain Project, a Swiss brain research initiative led by the academic institution Ecole Polytechnique Fédérale de Lausanne.

Called Blue Brain 5, the new supercomputer designed by HPE is based on the company's SGI 8600 system, a machine used to perform demanding HPC workloads.

With a budget of 18 million Swiss francs, about 13.6 million pounds sterling. Blue Brain 5 has 372 compute nodes to provide 1.06 petaflops of peak performance. It has 94 terabytes of memory, which equates to 23,000 laptops, and uses the Intel Xeon Gold 6140 and Xeon Phi 7230 processors as well as the Nvidia Tesla V100 graphics processors.

Add to the performance weight of the Mellanox InfiniBand dual-rail supercomputer and high-performance networks, as well as 4 petabytes of high-performance storage from DataDirect networks that provide aggregated bandwidth of more than 50 GB / s, badociated an 80 GB / s Infinite Memory Engine flash burst buffer. 19659002] Other technologies found in the HPC world have also made their way into Blue Brain 5, such as an easy-to-manage architecture and an energy-efficient liquid cooling system that avoids breaking down the heat. hot air in the supercomputer room.

All these performances will be used to better understand the mammalian brain, with the supercomputer aiming to help the Blue Brain Project model entire brain regions of a mouse by 2020.

With respect to mammalian brains, we can hope to improve knowledge of brain disorders from modeling and simulation, because these disorders are complex phenomena that extend from genes to ganglion and synapse circuits. [19659002"TheBlueBrainProjectScientificResultsheavilyonthecapabilitiesofthesupercomputer"saidFelixSchürmannco-directoroftheBlueBrainProject

"Modeling an individual neuron to Blue Brain today leads to about 20,000 ordinary differential equations – modeling entire regions of the brain, this quickly rises to $ 100 billion. Equations that must be solved simultaneously HPE helps us navigate the difficult technological landscape in supercomputers. "

While cloud computing is often presented as an easy and cost-effective way to access large amounts of power from calculation, the Blue Brain Project shows that there is still a lot of possibility for the supercomputer to solve the most complex problems that humans are facing.

Image credit: HPE

[ad_2]
Source link