Connect with us

Published

on

The world’s most powerful supercomputer, El Capitan, has been officially launched at the Lawrence Livermore National Laboratory (LLNL) in California. Built at a cost of $600 million, the system has been designed to manage highly classified national security tasks. The primary objective of the supercomputer is to ensure the security and reliability of the U.S. nuclear stockpile in the absence of underground testing, which has been prohibited since 1992. Research in high-energy-density physics, material discovery, nuclear data analysis, and weapons design will be conducted, along with other classified operations.

Performance and Capabilities

According to reports, El Capitan became the fastest supercomputer globally after achieving 1.742 exaFLOPS in the High-Performance Linpack (HPL) benchmark. The system has a peak performance of 2.746 exaFLOPS, making it the third machine ever to reach exascale computing speeds. The measurement, taken in floating-point operations per second (FLOPS), represents the ability of the supercomputer to perform one quintillion (10^18) calculations per second.

As reported by space.com, the second-fastest supercomputer, Frontier, located at Oak Ridge National Laboratory in Illinois, has recorded a standard performance of 1.353 exaFLOPS, with a peak of 2.056 exaFLOPS. El Capitan’s significant advancement marks a leap in computational capabilities within high-performance computing.

Technical Specifications

As reported by The Next Platform, El Capitan is powered by over 11 million processing and graphics cores distributed across 44,544 AMD MI300A accelerated processing units. These units incorporate AMD EPYC Genoa CPUs, AMD CDNA3 GPUs, and shared computing memory. Each processing unit includes 128 gigabytes of high-bandwidth memory, designed to optimise computational efficiency while minimising power consumption.

Development and Commissioning

Reports indicate that construction of El Capitan began in May 2023, with the system going online in November 2024. The official dedication took place on January 9, 2025. The supercomputer was commissioned by the U.S. Department of Energy’s CORAL-2 program as a successor to the Sierra supercomputer, which was deployed in 2018 and currently ranks 14th in the latest Top500 list of most powerful supercomputers.

With El Capitan’s full-scale deployment, advancements in national security research and computational science are expected to reach unprecedented levels.

Continue Reading

Science

AI Model Learns to Predict Human Gait for Smarter, Pre-Trained Exoskeleton Control

Published

on

By

Scientists at Georgia Tech have created an AI technique that pre-trains exoskeleton controllers using existing human motion datasets, removing the need for lengthy lab-based retraining. The system predicts joint behavior and assistance needs, enabling controllers that work as well as hand-tuned versions. This advance accelerates prototype development and could improve…

Continue Reading

Science

Scientists Build One of the Most Detailed Digital Simulations of the Mouse Cortex Using Japan’s Fugaku Supercomputer

Published

on

By

Researchers from the Allen Institute and Japan’s University of Electro-Communications have built one of the most detailed mouse cortex simulations ever created. Using Japan’s Fugaku supercomputer, the team modeled around 10 million neurons and 26 billion synapses, recreating realistic structure and activity. The virtual cortex offers a new platform for studying br…

Continue Reading

Science

UC San Diego Engineers Create Wearable Patch That Controls Robots Even in Chaotic Motion

Published

on

By

UC San Diego engineers have developed a soft, AI-enabled wearable patch that can interpret gestures with high accuracy even during vigorous or chaotic movement. The armband uses stretchable sensors, a custom deep-learning model, and on-chip processing to clean motion signals in real time. This breakthrough could enable intuitive robot control for rehabilitation, indus…

Continue Reading

Trending