All posts tagged: supercomputer

The UK Is Betting on a Billion-Dollar AI Supercomputer to Kick Its Addiction to US Tech

The UK Is Betting on a Billion-Dollar AI Supercomputer to Kick Its Addiction to US Tech

Published by skeptic

The UK government has laid out a $1.47 billion plan to shake its dependence on foreign-made artificial intelligence hardware. Under the measures, announced Monday, the UK will spend more than $1 billion on a national AI supercomputer. It will be stocked with $530 million worth of hardware, including $200 million that will go toward specialist inference chips for processing AI tasks. Priority will be given to up-and-coming British firms in the procurement process; the government pointed to Olix and Fractile, two UK startups developing new styles of inference chip, as potential beneficiaries. British researchers and startups are expected to be able to use the supercomputer starting in 2030. The new measures are part of a broader effort by the UK government to minimize dependence on foreign powers for access to AI products and services—a move made more urgent by the apparent souring of the relationship between the US and its European counterparts. The European Union outlined a similar “tech sovereignty” proposal last week. This year, European leaders have found themselves in confrontation with the Trump …

EuroHPC backs €290M AI supercomputer for IT4LIA AI factory

EuroHPC backs €290M AI supercomputer for IT4LIA AI factory

Published by skeptic

A major public investment is being directed into Italy’s next phase of high-performance AI infrastructure, with a new AI supercomputer commissioned for the IT4LIA AI factory in Bologna. The initiative is being led by the European High Performance Computing Joint Undertaking (EuroHPC JU) in partnership with E4 Computer Engineering and Dell Technologies. The system forms part of a €290m programme covering procurement through to long-term operation. Funding is split evenly, with support from the Digital Europe Programme (DEP) and co-financing from Italy’s Ministry of University and Research. When deployed, the platform, delivering more than 160 exaflops of peak AI inference performance, will anchor Italy’s AI factory within a wider European network, delivering large-scale AI compute capacity and expanding access for startups, SMEs, and researchers working on data-intensive applications. Anders Jensen, Executive Director of EuroHPC JU, commented: “This signature marks another important milestone for Europe’s HPC and AI ambitions. “With the IT4LIA system, we are making a significant investment in cutting-edge infrastructure that will empower startups, SMEs, and the research community to develop and deploy advanced AI …

EuroHPC and Bull collab to deploy AI supercomputer at Mimer AIF

EuroHPC and Bull collab to deploy AI supercomputer at Mimer AIF

Published by skeptic

The EuroHPC Joint Undertaking (EuroHPC JU) has signed a procurement contract with Bull to deploy a new AI supercomputer in Linköping, Sweden. The system will underpin the next phase of the Mimer AI Factory (AIF) initiative, expanding Europe’s AI infrastructure. The AI supercomputer will be hosted and operated by the National Academic Infrastructure of Supercomputing in Sweden (NAISS) at Linköping University. It will support the Mimer AIF in combining cloud-enabled supercomputing, large-scale sensitive data storage, and advanced software layers to accelerate AI development across multiple industries. The immediate result is a significant increase in computing capacity and service capability within the Mimer AIF ecosystem. This expansion is expected to strengthen Sweden’s role in Europe’s AI Factory network while enabling more advanced applications across healthcare, materials science, and autonomous systems. What does the Mimer AIF expansion actually deliver? The new AI supercomputer is designed to move Mimer AIF from a support platform into a high-capacity AI development hub. It integrates compute, storage, and specialised expertise into a single environment aimed at production-level AI workloads. This includes …

ARCHER2 supercomputer generates £4.2bn for UK economy

ARCHER2 supercomputer generates £4.2bn for UK economy

Published by skeptic

Britain’s national high-performance computing system delivers more than an eightfold return on public investment while underpinning global research output. The UK’s flagship computing system, the ARCHER2 supercomputer, has produced an estimated £4.2bn in economic value since its launch, according to a newly released independent assessment. The findings suggest the system, which is hosted at the EPCC at the University of Edinburgh, has delivered a return of more than eight times the public funding invested in its development and operation. The analysis, commissioned by UK Research and Innovation (UKRI) and conducted by London Economics, evaluates both the economic and scientific contributions of ARCHER2 over its first five years in service. Economic impact driven by research and innovation The report estimates that for every £1 of public funding allocated to the ARCHER2 supercomputer, approximately £8.30 has been returned to the UK economy. This calculation factors in around £100m in capital investment to build the system alongside roughly £400m in associated research funding. Most of the economic benefit – around £3.7bn – originates from research and development activity …

UK unveils £45m Sunrise AI supercomputer to accelerate fusion

UK unveils £45m Sunrise AI supercomputer to accelerate fusion

Published by skeptic

An industry-transforming system aims to position Britain at the forefront of AI-driven clean energy innovation. The UK Government has committed £45m to develop a new 1.4MW AI supercomputer designed to support fusion energy research, marking a significant step in integrating artificial intelligence with advanced scientific infrastructure. The system, named Sunrise, will be based at the UK Atomic Energy Authority (UKAEA) Culham Campus in Oxfordshire and is expected to become operational by mid-2026. The project forms part of a broader effort to establish the UK’s first “AI Growth Zone,” combining high-performance computing with national energy research priorities. Officials position the initiative as both a technological investment and a strategic move to strengthen domestic capabilities in clean energy and AI. Dr Rob Akers, UKAEA’s Director for Computing Programmes, explained: “UKAEA is taking lessons from the Apollo programme: we learn fastest when we can test, iterate, and improve safely in the virtual world before we commit to our real-world mission. “Sunrise will bring that capability to fusion by combining high-fidelity simulation with physics-informed AI to develop predictive digital twins that reduce the cost, risk and time …

Nvidia’s DGX Station is a desktop supercomputer that runs trillion-parameter AI models without the cloud

Nvidia’s DGX Station is a desktop supercomputer that runs trillion-parameter AI models without the cloud

Published by skeptic

Nvidia on Monday unveiled a deskside supercomputer powerful enough to run AI models with up to one trillion parameters — roughly the scale of GPT-4 — without touching the cloud. The machine, called the DGX Station, packs 748 gigabytes of coherent memory and 20 petaflops of compute into a box that sits next to a monitor, and it may be the most significant personal computing product since the original Mac Pro convinced creative professionals to abandon workstations. The announcement, made at the company’s annual GTC conference in San Jose, lands at a moment when the AI industry is grappling with a fundamental tension: the most powerful models in the world require enormous data center infrastructure, but the developers and enterprises building on those models increasingly want to keep their data, their agents, and their intellectual property local. The DGX Station is Nvidia’s answer — a six-figure machine that collapses the distance between AI’s frontier and a single engineer’s desk. What 20 petaflops on your desktop actually means The DGX Station is built around the new …

DIY ,500 Supercomputer Replaces Cloud and AI Costs

DIY $8,500 Supercomputer Replaces Cloud and AI Costs

Published by skeptic

Jay has unveiled Zeus, a custom-built supercomputer designed for $8,500 to tackle the rising costs of cloud services. Equipped with an AMD Ryzen 9 CPU, 128GB of RAM and an Nvidia 5090 GPU, Zeus is optimized for tasks such as data scraping, email verification and AI model training. Its modular setup, powered by Unraid OS and Docker containers, enables businesses to customize workflows and reduce reliance on external infrastructure. Explore how Zeus minimizes cloud expenses while supporting tasks like website management and local data storage. Gain insight into its scalable architecture and how it facilitates AI-driven operations. Learn practical steps for adapting similar systems to meet specific business requirements. Zeus: Cost-Effective AI Solution TL;DR Key Takeaways : Zeus, a custom-built supercomputer costing $8,500, offers a cost-effective alternative to cloud services by hosting AI models, automating workflows and managing data locally, reducing operational expenses by up to 80%. Equipped with high-performance components like an AMD Ryzen 9 CPU, 128GB RAM, 14TB storage and Nvidia 5090 GPU, Zeus is optimized for demanding tasks such as AI model …

Why Sierra the Supercomputer Had to Die

Why Sierra the Supercomputer Had to Die

Published by skeptic

Supercomputers can be measured in several ways, but the vital statistic is their ability to perform floating-point operations per second, or flops. Flopping as fast as possible is what makes you successful. At her peak, Sierra could hit 94.64 petaflops—94.64 quadrillion floating-point operations—per second. El Capitan, at 1.809 exaflops, is about 19 times faster. In late 2025, he was officially declared the world’s fastest supercomputer. Sierra’s juice, Neely says, was no longer worth the squeeze. There was no big red button, no giant lever, that turned Sierra off. Someone could’ve just cut the cords, sure, but that’s not the recommended procedure. First, Sierra’s user scientists were warned, via email, to save their work. Then a DNR was formally instituted—no new parts. The decommissioning proceeded in phases, starting with the compute nodes and the rack switches—management nodes are last, since they’re needed until the very end. The process involves running scripts that, digitally, shut the computer down, and then hard power switches are flipped off too. There’s also a dehydration. When she was alive, Sierra could …

£36m DAWN supercomputer upgrade to supercharge UK AI

£36m DAWN supercomputer upgrade to supercharge UK AI

Published by skeptic

British science is gearing up for a major leap forward after the government confirmed a £36m investment to dramatically expand the power of one of the country’s most important computing hubs. The funding will increase capacity at the University of Cambridge’s AI facilities by six times, strengthening the UK’s ability to develop next-generation artificial intelligence and data-driven technologies. At the centre of this expansion is the DAWN supercomputer, a system that has already become a cornerstone of national AI research. The upgrade is expected to come online as early as spring 2026, unlocking new possibilities for researchers, start-ups, and public-sector innovators alike. Professor Sir John Aston, Pro-Vice-Chancellor for Research, University of Cambridge, explained: “This investment marks an important milestone for the UK’s AI Research Resource, expanding the power of Cambridge’s DAWN supercomputer and strengthening our national computing ecosystem. “It will give researchers, clinicians and innovators the tools they need to drive breakthroughs that improve public services. “The University of Cambridge is proud to work with industry leaders such as Dell to ensure world-class computing is available to …

Scientists Preparing to Simulate Human Brain on Supercomputer

Scientists Preparing to Simulate Human Brain on Supercomputer

Published by José Betancourt

In 2024, researchers completed the first-ever map of the circuitry of a fruit fly’s brain. Despite its diminutive size, the organ packs almost 500 feet of wiring and 54.5 million synapses into the size of a grain of sand — an astonishing feat of computational neurology research that allows scientists to better understand how signals travel throughout the brain. And thanks to significant advances of some of the world’s most capable supercomputers, researchers at the Jülich Research Centre in Germany are now aiming their sights at a far more ambitious goal: a simulation at the scale of the entire human brain. Previous attempts, dating back a decade, like the Human Brain Project, fell largely flat, despite considerable government funding. But as New Scientist reports, the Jülich researchers think they can push things forward. The idea is to bring together several models of smaller regions of the brain with a supercomputer to run simulations of billions of firing neurons. The team, which is being led by Jülich neurophysics professor Markus Diesmann, will leverage the Joint Undertaking …