El Capitan, the record-breaking computer built by HPE and housed at the Lawrence Livermore National Laboratory in California, has actually claimed the title of the country’s fastest computer, according to the latest TOP500 positions. The renowned record, tracking multiprocessor milestones since 1993, underlines El Capitan’s pioneering achievement.
” El Capitan marks another significant milestone in exascale supercomputing, bringing huge performance, energy efficiency, and the capabilities to promote AI-driven medical discovery and create amazing breakthroughs to improve national security and unlock fresh opportunities in green energy”, Trish Damkroger, senior vice president and general manager, HPC &, AI Infrastructure Solutions, said in a press release.
El Capitan forces state clinical research, including nuclear tests simulations
El Capitan was clocked at a Linpack report of 1.742 EFlop/s, a standard indicator of how fast a sophisticated sequence of linear equations can be solved.
El Capitan supports the function of the U. S. National Nuclear Security Administration, including research on atomic weapons and barriers. The laboratory even uses the system for clinical studies on energy security, climate change, energy grid development, and drug discovery.
The computer uses AMD 4th-generation EPYC chips, the Cray Slingshot 11 system for data move, and a specialty backup solution. The country’s number one computer has 11, 039, 616 combined CPU and GPU components. It runs at 58.89 gigaflops achievement per watt.
El Capitan’s introduction adds to the capability development required to keep our stockpile safe without triggering new, violent nuclear testing, according to Jill Hruby, the undersecretary for nuclear security and administrator of the Department of Energy.
In May 2024, El Capitan appeared on the TOP500 record at number 46.
SEE: At this week’s global computation conference, Dell provided new facilities for enterprise AI applications.
Top 5 Mainframes of 2024
Behind El Capitan, here are the most prominent computers in the world, according to the TOP500 record:
Oak Ridge National Laboratory in Tennessee’s Frontier method
El Capitan could n’t keep Frontier from capturing the top spot, despite Frontier increasing its exaflops to 1.353 Eflop/s on the November list. Frontier uses 9, 066, 176 components and Cray’s Hammer 11 system. It is used to analyze materials, power, nuclear power, and biology, among other medical research.
The Aurora program at the Illinois-based Argonne Leadership Computing Service
With Intel Xeon CPU Max Series Processors and Intel Data Center GPU Max Series Accelerators, Aurora is constructed using HPE’s Cray EX — Intel Exascale Compute tip. It maintained its standard score from earlier this year, at 1.012 Exaflop/s. The laboratory uses the computer for technological research, including weather, materials, fusion power, and energy storage.
Microsoft Azure’s Eagle method
A cloud-based computer swarm is provided by Microsoft Azure. Eagle uses 14, 400 NVIDIA H100 GPUs and Intel Xeon Sapphire Rapids computers to achieve a standard rating of 561.2 PFlop/s.
The HPC6 program at Eni S. p. A facility in Ferrera Erbognone, Italy
HPC6 joined the top five this month, bringing 477.90 PFlop/s from its home with the Eni power company in Italy. The fastest system in Europe is then presently be HPC6. TOP500 records HPC6 uses the exact layout as Frontier.
The top 10 are powered by AMD and Intel chips.
TOP500 displays AMD and Intel are the best options for the world’s most powerful microprocessors. Five of the top 10 employ AMD chips, while three apply Intel. The remaining two apply NVIDIA or a specialty ARM-based computer. The Slingshot-11 connection seems to be the most popular option for communication, used by seven of 10.
