Frontier, the ex HPE Cray supercomputer operated by the Department of Energy (DoE) at Oak Ridge National Laboratory (ORNL), remains by far the fastest supercomputer on the planet, according to the Top 500.
With a high-performance Linpack (HPL) benchmark of 1,102 Exaflops per second, Frontier is the only ex-scale supercomputer and is three times faster than its closest competitor, Fugaku, at the Center for Machine Science. RIKEN calculation (R-CCS) in Kobe, Japan, according to Top500. An exaflop is one million calculations or floating point operations (flops) per second.
Frontier, an HPE Cray EX system, is equipped with 8,730,112 cores and combines a 3rd Gen AMD EPYC CPU optimized for HPC and AI, with an AMD Instinct 250X accelerator and Slingshot-10 connectivity.
Frontier took the top spot in the Top 500 rankings in June, achieving a similar 1.1 exaflop performance. It has the highest theoretical 2 exaflops and is 10 times more powerful than ORNL’s Summit system.
Also: White House Passes AI ‘Bill of Rights’ to Try to Allay Your Anxieties
Frontier is also part of the DoE Exacscale Computer Project. It’s housed in 74 massive HPE Cray EX supercomputer cabinets, including more than 9,400 AMD-supplied nodes and 90 miles of network cabling, according to ORNL. It also has 37,000 GPUs.
As ORNL explainsModeling cancer cells, supernovae, coronaviruses, or the atomic structure of elements requires the 64-bit precision that conventional supercomputers have, while machine algorithms learning requires less precession, such as 32-, 24-, or 16-bit precision and thus may lose the advantage of the GPU.
Frontier’s Orion storage system holds 700 Petabytes of data; Each cabinet weighs 8,000 pounds, and the system is cooled by 6,000 gallons of water pumped per minute by four 350-horsepower pumps. The pump can fill an Olympic-sized swimming pool in 30 minutes. The Top 500 press release said: “Frontier is clearly the winner of the exascale race, and it will require a lot of effort and innovation to knock it off the top spot.”
The fastest supercomputer in Europe, in third place in the Top 500, is the upgraded LUMI system based in Finland, which is also the HPE Cray EX system. It is installed at the EuroHPC center at CSC’s data center in Kajaani, Finland and has an efficiency of 309.1 Pflop/s.
As Top 500 notes, European Joint Commitment to High Performance Computing (EuroHPC JU) is pooling European resources to develop leading Exascale supercomputers for big data processing.
A new entrant is Leonardo, installed at the EuroHPC site at Cineca, in Italy. It’s a French-made Atos BullSequana XH2000 system with Xeon Platinum 8358 32C 2.6GHz as the main processor, a 40 GB NVIDIA A100 SXM4 as the accelerator, and the NVIDIA HDR100 Infiniband Quad-rail as the interconnect. It achieved a Linpack performance of 174.7 Pflop/s.
Atos earlier this year XH300 launcheda more powerful system than the XH200 with CPUs, GPUs and accelerators from AMD, Intel and Nvidia, as well as microprocessor technology from the French company The Silicon Pearl (SiPearl).
Old Summit, a system built by IBM, is in fifth place with a performance of 148.8 Pflop/s.
Sierra, Summit’s sister at Lawrence Livermore National Laboratory, California, was in sixth place with a score of 94.6 Pflop/s.
China’s fastest supercomputer, Sunway TaihuLight, is The world’s top performing supercomputer in 2017, but now ranks seventh with a performance of 93 Pflops/s. Tianhe-2A (Milky Way-2A) used to be the second fastest in 2017 but is now in 10th place with a performance of 61.4 Pflop/s. They did overshadowed by Summit and Sierra in 2018.
While China’s fastest supercomputers slide in the HPL Top 500 rankings, China still dominates the Top 500 list alongside the US. The US has 126 machines in the Top 500, while China has 162 machines in this ranking. However, the number of Chinese systems fell from 173 in the June 2022 ranking. Europe as a whole accounted for 131 of the Top 500, up from 118 in June.
