Nvidia has partnered with the College of Florida to upgrade the university’s HiPerGator supercomputer. Set to go are living in early 2021, the challenge will make HiPerGator the speediest AI supercomputer in academia with 700 petaflops of functionality, according to Nvidia and UF.
HiPerGator three. will use a cluster comprised of one hundred forty Nvidia DGX A100 nodes, which are $two hundred,000 supercomputing AI techniques that each individual have eight A100 GPUs. The DGX A100, offering 320 GB of RAM for training enormous AI datasets, is able of 5 petaflops of functionality.
The cluster of one hundred forty DGX A100 nodes, called a DGX SuperPOD, will therefore strike 700 petaflops, Nvidia and the university reported. HiPerGator three. will also contain 4 petabytes of substantial-functionality storage and tap Nvidia Mellanox HDR 200Gb/s InfiniBand networking.
“The projected 700 petaflops are certainly up there,” reported Nick McQuire, senior vice president and head of AI and enterprise research at CCS Perception. “It is section of an arms race we are looking at at a national amount and within just academia to ensure universities remain aggressive globally.”
UF is the 1st educational institute to use the DGX A100 technology, which was declared in Might.
The speediest supercomputer on the Tremendous Laptop Top 500 record from June is a program at RIKEN Heart for Computational Science in Japan called Supercomputer Fugaku, which hits 415 petaflops of functionality. For unique tasks, Fugaku peaks at one,000 petaflops, or one exaflop.
Nvidia’s new A100 architecture will assistance electrical power the College of ‘Florida’s HiPerGator three. supercomputer.
If Nvidia’s statements bear out, UF’s approaching program will be comparable to Fugaku in conditions of electrical power.
It is section of an arms race we are looking at at a national amount and within just academia to ensure universities remain aggressive globally. Nick McQuireAnalyst, CCS Perception
Supercomputing competitiveness shifts
Up until now, most of the substantial-profile competitiveness in AI supercomputing has centered on program speed and been funded at the national govt amount, these kinds of as with Fugaku, McQuire reported.
The Top 500 record does contain a lot of supercomputers at universities, together with the College of Texas, Nagoya College in Japan and Mississippi Condition College. Nvidia technology assists electrical power a lot of of the supercomputers on the record, as do HPE and IBM. The a few tech giants dominate the supercomputing industry alongside suppliers these kinds of as Fujitsu, Dell, Lenovo, Atos and Penguin Computing.
The UF challenge is being funded by a $25 million donation from UF alumnus and Nvidia co-founder Chris Malachowsky, along with $25 million in components, program, services and training from Nvidia. UF also pledged $twenty million of its have cash.