November 2020

The most energy-efficient system on the Green500 is the new NVIDIA DGX SuperPOD in the US. It achieved 26.2 gigaflops/watt power-efficiency during its 2.4 HPL performance run and is listed at position 170 in the TOP500.

Next on the list is the previous Green500 champ, MN-3. Although it improved its score from 21.1 to 26.0 gigaflops/watt, it slips into the number two position. The system uses the MN-Core chip, an accelerator optimized for matrix arithmetic. It is ranked number 330 in the TOP500.

In the number three Green500 is the Atos-built JUWELS Booster Module installed at Forschungszentrum Jülich (FZJ) in Germany. It achieves 25.0 gigaflops/watt and is ranked seventh in the TOP500.

In fourth position is Spartan-2, another Atos-built machine. It achieves 24.3 gigaflops/watt on HPL and is ranked at position 148 on the TOP500 list.

The fifth-ranked system on the Green500 is Selene, with an efficiency of 24.0 gigaflops/watt. It also occupies the number five spot on the TOP500.

With the exception of the MN-3 system, the remaining top five Green500 systems are using the new NVIDIA A100 GPU as an accelerator. All four of these systems use AMD EPYC as their main CPU.

Of the top 40 systems on the Green500, 37 leverage accelerators, 2 use A64FX vector-processors, and one (TaihuLight) a Sunway many-core processor.

Extrapolating the power efficiency value of 26.2 gigaflops/watt of the NVIDIA DGX SuperPOD out linearly to an exaflop would result in a power consumption of 38 MW (ignoring additional hardware needed for scaling).

Green500 Data

Rank TOP500 Rank System Cores Rmax (PFlop/s) Power (kW) Energy Efficiency (GFlops/watts)
1 170 NVIDIA DGX SuperPOD - NVIDIA DGX A100, AMD EPYC 7742 64C 2.25GHz, NVIDIA A100, Mellanox HDR Infiniband, Nvidia
NVIDIA Corporation
United States
19,840 2.36 90 26.195
2 330 MN-3 - MN-Core Server, Xeon Platinum 8260M 24C 2.4GHz, Preferred Networks MN-Core, MN-Core DirectConnect, Preferred Networks
Preferred Networks
Japan
1,664 1.65 65 26.039
3 7 JUWELS Booster Module - Bull Sequana XH2000 , AMD EPYC 7402 24C 2.8GHz, NVIDIA A100, Mellanox HDR InfiniBand/ParTec ParaStation ClusterSuite, EVIDEN
Forschungszentrum Juelich (FZJ)
Germany
449,280 44.12 1,764 25.008
4 146 Spartan2 - Bull Sequana XH2000 , AMD EPYC 7402 24C 2.8GHz, NVIDIA A100, Mellanox HDR Infiniband, EVIDEN
Atos
France
23,040 2.57 106 24.262
5 5 Selene - NVIDIA DGX A100, AMD EPYC 7742 64C 2.25GHz, NVIDIA A100, Mellanox HDR Infiniband, Nvidia
NVIDIA Corporation
United States
555,520 63.46 2,646 23.983
6 239 A64FX prototype - Fujitsu A64FX, Fujitsu A64FX 48C 2GHz, Tofu interconnect D, Fujitsu
Fujitsu Numazu Plant
Japan
36,864 2.00 118 16.876
7 29 AiMOS - IBM Power System AC922, IBM POWER9 20C 3.45GHz, NVIDIA Volta GV100, Dual-rail Mellanox EDR Infiniband, IBM
Rensselaer Polytechnic Institute Center for Computational Innovations (CCI)
United States
130,000 8.34 512 16.285
8 8 HPC5 - PowerEdge C4140, Xeon Gold 6252 24C 2.1GHz, NVIDIA Tesla V100, Mellanox HDR Infiniband, DELL
Eni S.p.A.
Italy
669,760 35.45 2,252 15.740
9 458 Satori - IBM Power System AC922, IBM POWER9 20C 2.4GHz, Infiniband EDR, NVIDIA Tesla V100 SXM2, IBM
MIT/MGHPCC Holyoke, MA
United States
23,040 1.46 94 15.574
10 1 Supercomputer Fugaku - Supercomputer Fugaku, A64FX 48C 2.2GHz, Tofu interconnect D, Fujitsu
RIKEN Center for Computational Science
Japan
7,630,848 442.01 29,899 15.418