MORRISVILLE – Supercomputer technology ranging from liquid cooling to banks of high-end servers from Lenovo powered by new generation Intel processors are delivering power that the companies say is capable of making advances in earthquake forecasting to new research into disease forecasting.
The world’s top seller of traditional PCs has teamed up with Intel to build Harvard University’s newest supercomputer that’s capable of improving even exascale computing.
Exascale means the power to make a billion billion, or quintillion, calculations per second.
That’s what some scientists calculate is the speed of the human brain, And hitting exascale “represents a thousandfold increase over the first petascale computer that came into operation in 2008,” notes Wikipedia.
Lenovo, which is the world’s biggest supercomputer manufacturer, bases its supercomputer and data center business group based in Research Triangle Park. And it says “direct-to-node water-cooling” is a better way than conventional airflow technology to keep the banks of servers running.
But the Harvard project may not produce the world’s fastest calculations.
The U.S. Department of Energy is working with supercomputer firm Cray to build a 1,5 exaflop supercomouter that’s due online in 2021 and will be the world’s fastest. But in the meantime Lenovo cranks out powerful machines of its own, including the Harvard Faculty of Arts and Sciences Research Computing project.
Lenovo, Intel to collaborate on driving artificial intelligence to improve supercomputers
Monday’s news follows an announcement in August when Lenovo and Intel disclosed plans to drive improvements in high-performance computing and artificial intelligence.
The tech giants said their “multiyear collaboration” will have their firms looking to enable computers “accelerate solutions for the world’s most challenging problems.”
“Science is all about iteration and repeatability. But iteration is a luxury that is not always possible in the field of university research because you are often working against the clock to meet a deadline,” said Scott Yockel, director of research computing at Harvard University’s Faculty of Arts and Sciences, in Monday’s announcement.
“With the increased compute performance and faster processing of the Cannon cluster, our researchers now have the opportunity to try something in their data experiment, fail, and try again. Allowing failure to be an option makes our researchers more competitive.”
Lenovo’s ThinkSystem SD650 NeXtScale servers with Neptun liquid cooling technology and Intel Xeon Platinum 8268 processors power the Harvard cluster, which is based in Holyoke, MA.
“Each Cannon [the name of the storage system] node is now several times faster than any previous cluster node, with jobs like geophysics models of the Earth performing 3-4 times faster than the previous system,” Lenovo and Intel say.
