Frontier retains top spot among world’s fastest supercomputers

Frontier held onto its No. 1 ranking in the 63rd edition of the TOP500, but with the second place Aurora system breaking the exascale machine barrier, the end of Frontier’s reign could be in sight.

The Frontier system at Oak Ridge National Laboratory (ORNL) in Tenn., maintained its leading position with an HPL score of 1.206 EFlop/s. With a total of 8,699,904 combined CPU and GPU cores, the Frontier system has an HPE Cray EX architecture that combines 3^rd Gen AMD EPYC CPUs optimized for HPC and AI with AMB Instinct MI250X accelerators. The system relies on Cray’s Slingshot 11 network for data transfer, and the machine has a power efficiency rating of 52.93 GFlops/Watt –which also puts Frontier at the No. 13 spot on the GREEN500.

Staying in line with the last list, the Aurora system at the Argonne Leadership Computing Facility in Ill. is ranked second on the TOP500. Aurora is also now the second machine to break the exascale barrier with an HPL score of 1.012 EFlop/s – an improvement over the 585.34 PFlop/s score for the last edition. The Aurora system is based on HPE Cray EX- Intel Exascale Computer Blade and uses Intel Xeon CPU Max series processors, Intel Data Center GPU Max Series accelerators, and a Slingshot-11 interconnect. TOP500 notes that Aurora achieved this rank while currently being decommissioned and not fully complete.

The systems rounding out the top five also remained consistent. A system called Eagle disrupted the list in late 2023, but it maintained its No. 3 position in this ranking. Installed on the Microsoft Azure Cloud in the U.S. The Eagle system continues to be the highest-ranking cloud system on the TOP500. The Eagle’s Microsoft NDv5 system has an HPL score of 561.2 PFlop/s and is based on Intel Xeon Platinum 8480C processors and Nvidia H100 accelerators.

The Supercomputer Fugaku and LUMI systems also retained their No. 4 and No. 5 positions, respectively. Based in Kobe, Japan, Fugaku has an HPL score of 442 PFlop/s and continues to be the highest-ranked system outside of the U.S. The LUMI system at EuroHPC/CSC in Finland – the largest system in Europe – also stayed put in its No. 5 spot with an HPL core of 380 PFlop/s.

One newcomer to the list, the Alps machine from the Swiss National Supercomputing Centre (CSCS) in Switzerland, achieved an HPL score of 270 PFlop/s, landing at No. 6 in this edition’s ranking. Sierra, the system installed on at the Lawrence Livermore National Laboratory in Calif., fell off the list this time, previously ranked No. 10 in November 2023.

Here is a breakdown of specific details for the 10 overall fastest supercomputer systems on the TOP500 list for May 2024:

#1: Frontier

This HPE Cray EX system is the first U.S. system with a performance exceeding one Exaflop/s. It is installed at the ORNL in Tenn., where it is operated for the Department of Energy (DOE). 

• Cores: 8,699,904
• Rmax (PFLOPS): 1,206.00
• Rpeak (PFLOPS): 1,714.81
• Power (kW): 22,786

#2: Aurora

The Aurora system is installed at the Argonne Leadership Computing Facility, Illinois, USA, where it is also operated for the DOE and holds a preliminary HPL score of 1.012 Exaflop/s. 

• Cores: 9,264,128
• Rmax (PFLOPS): 1,012.00
• Rpeak (PFLOPS): 1,980.01
• Power (kW): 38,698

#3: Eagle

The No. 3 system is installed by Microsoft in its Azure cloud. This Microsoft NDv5 system is based on Xeon Platinum 8480C processors and Nvidia H100 accelerators and achieved an HPL score of 561 Pflop/s.

• Cores: 2,073,600
• Rmax (PFLOPS): 561.20
• Rpeak (PFLOPS): 846.84

#4: Supercomputer Fugaku

This system is installed at the RIKEN Center for Computational Science (R-CCS) in Kobe, Japan. It has 7,630,848 cores which allowed it to achieve an HPL benchmark score of 442 Pflop/s.

• Cores: 7,630,848
• Rmax (PFLOPS): 442.01
• Rpeak (PFLOPS): 537.21
• Power (kW): 29,899

#5: LUMI

Located in CSC’s data center in Kajaani, Finland, the European High-Performance Computing Joint Undertaking (EuroHPC JU) is pooling European resources to develop top-of-the-range Exascale supercomputers for processing big data.

• Cores: 2,752,704
• Rmax (PFLOPS): 379.70
• Rpeak (PFLOPS): 531.51
• Power (kW): 7,107

#6: Alps (new to the list)

This system is an HPE Cray EX254n system with Nvidia Grace 72C and Nvidia GH200 Superchip and a Slingshot-11 interconnect, achieving 270 PFlop/s.

• Cores: 1, 305,600
• Rmax (PFLOPS): 270.00
• Rpeak (PFLOPS): 353.75
• Power (kW): 5,194

#7: Leonardo (previously #6)

The Leonardo system is installed at another EuroHPC site in CINECA, Italy. It is an Atos BullSequana XH2000 system with Xeon Platinum 8358 32C 2.6GHz as main processors, NVIDIA A100 SXM4 40 GB as accelerators, and Quad-rail NVIDIA HDR100 Infiniband as interconnect.

• Cores: 1,824,768
• Rmax (PFLOPS): 241.20
• Rpeak (PFLOPS): 306.31
• Power (kW): 7,494

#8: MareNostrum 5 ACC

The MareNostrum 5 ACC system was remeasured and jumped in the ranking over the Summit system. It is now at No. 8 and installed at the EuroHPC/Barcelona Supercomputing Center in Spain.

• Cores: 663,040
• Rmax (PFLOPS): 175.30
• Rpeak (PFLOPS): 249.44
• Power (kW): 4,159

#9: Summit (previously #7)

Housed at the ORNL in Tenn., the IBM-built Summit system has 4,356 nodes, each one housing two POWER9 CPUs with 22 cores each and six NVIDIA Tesla V100 GPUs each with 80 streaming multiprocessors (SM). 

• Cores: 2,414,592
• Rmax (PFLOPS): 148.60
• Rpeak (PFLOPS): 200.79
• Power (kW): 10,096

#10: Eos NVIDIA DGX SuperPOD (previously #9)

This system is based on the NVIDIA DGX H100 with Xeon Platinum 8480C processors, NVIDIA H100 accelerators, and Infiniband NDR400 and it achieves 121.4 PFlop/s.

• Cores: 485,888
• Rmax (PFLOPS): 121.40
• Rpeak (PFLOPS): 188.65