Hewlett Packard Enterprise says it’s been selected to build a supercomputer designed to simulate the inner workings of the mouse brain by 2020.
The computer, known as Blue Brain 5, will become the platform for the Blue Brain Project, a Swiss-led campaign to model and simulate the mammalian brain. The project — which is under the supervision of the Ecole Polytechnique Fédérale de Lausanne, or EPFL — meshes with international neuroscience efforts such as Europe’s Human Brain Project and the U.S. BRAIN Initiative.
“The Blue Brain Project’s scientific mission is critically dependent on our supercomputing capabilities,” project co-director Felix Schürmann said today in a news release announcing the collaboration with Hewlett Packard Enterprise.
“Modeling an individual neuron at Blue Brain today leads to around 20,000 ordinary differential equations – when modeling entire brain regions, this quickly raises to 100 billion equations that have to be solved concurrently,” Schürmann said. “HPE helps us to navigate the challenging technology landscape in supercomputing.”
HPE was awarded an initial contract for the project at the end of 2017, and follow-up work could bring the total value of the award to more than $18 million.
The supercomputer is based on the HPE SGI 8600 System and tailored to the objective of modeling entire regions of the mouse brain, with a focus on the thalamus and neocortex.
Getting a better sense of how the mouse brain works should provide hints about the more complex workings of the human brain, and reveal how diseases disrupt neurological function.
“Our mission is to create technologies that improve our quality of life, including powering technologies for the healthcare industry to deliver targeted treatments and save lives,” said Antonio Neri, HPE’s president and CEO. “Through our relationship with the Blue Brain Project, HPE is bringing advanced supercomputing and bespoke applications to empower new research that can have transformative benefits for the neuroscientific community and society at large.”
The Blue Brain Project’s HPE SGI 8600 supercomputer makes use of 372 compute nodes, delivering 1.06 petaflops of peak performance.
HPE says the system is equipped with 94 terabytes of memory – a memory equivalent of 23,000 laptops – and runs Intel Xeon Gold 6140 and Intel Xeon Phi 7230 processors as well as NVIDIA Tesla V100 graphic processors.
Brace yourself for more geekspeak: The system uses single and dual-rail Mellanox InfiniBand high-performance networks and has 4 petabytes of high-performance storage from DataDirect Networks. That configuration should deliver more than 50 GB/s aggregated bandwidth, associated with an 80 GB/s Infinite Memory Engine flash-based burst buffer.
The setup, installed at the Swiss National Supercomputing Center in Lugano, incorporates an energy-efficient liquid cooling system that won’t push heated air into the data center.
In other news, the U.S. Department of Energy’s IBM-built Summit supercomputer has captured the No. 1 spot on the global TOP500 list of supercomputers, with a processing speed of 122.3 petaflops on the benchmark High Performance Linpack, or HPL. The Summit computer was unveiled last month.
China’s Sunway TaihuLight dropped to No. 2 after leading the list for two years with an HPL performance of 93 petaflops. The Energy Department’s Sierra supercomputer is No. 3 with 71.6 petaflops.