Connect with us

Business

Isambard-AI, the UK’s Most Powerful AI Supercomputer, Goes Live

Published

on





Isambard-AI, the UK’s Most Powerful AI Supercomputer, Goes Live | NVIDIA Blog






















The University of Bristol’s Isambard-AI, powered by NVIDIA Grace Hopper Superchips, delivers 21 exaflops of AI performance, making it the fastest system in the U.K. and among the most energy-efficient globally.

The U.K. has officially joined the premier league of global AI infrastructure — and it’s not starting small.

At a ribbon-cutting ceremony at the Bristol Centre for Supercomputing (BriCS), leaders today unveiled Isambard-AI, the most powerful AI supercomputer ever built in the U.K.

U.K. Secretary of State Peter Kyle was joined by leaders from across academia, industry and government, including Simon McIntosh-Smith, director of the Bristol Centre for Supercomputing, and Neil MacDonald, executive vice president and general manager  for Hewlett Packard Enterprise’s server business.

And as we press this switch to activate the UK’s most powerful supercomputer, we are embarking on Britain’s super future where AI contributes towards the delivery of better public services, greater public prosperity, deeper scientific discovery and stronger national security,” Kyle said.

 The numbers back up Kyle’s statement:

  • 21 exaflops of AI performance
  • 5,448 NVIDIA GH200 Grace Hopper Superchips
  • Set to rank 11th worldwide on the latest TOP500 list of world’s fastest supercomputers
  • More than 10x faster than the next-fastest supercomputer in the U.K. More computing power than all other U.K. supercomputers combined
  • Ranked fourth globally for energy efficiency
Engineering teams working collaboratively to maneuver and install components, illustrating the complex and coordinated effort behind the build.


Engineering teams working collaboratively to maneuver and install components within the supercomputer’s data hall, illustrating the complex and coordinated effort behind the build.

Why It Matters

Isambard-AI gives U.K. researchers and businesses a once-in-a-generation leap in computing power. It’s a platform to accelerate breakthroughs in:

  • AI-driven drug discovery
  • Advanced climate modeling
  • Materials science
  • Large language models (LLMs) tuned to U.K. languages and laws

Backed by £225 million in government investment and built with NVIDIA, HPE, the University of Bristol and others, Isambard-AI signals a clear ambition: to lead in AI, the U.K. must lead in compute.

Simon McIntosh-Smith, director of the Bristol Centre for Supercomputing, with Simon Appleby, AI business lead at HPE.

Simon McIntosh-Smith, director of the Bristol Centre for Supercomputing, with Simon Appleby, AI business lead at HPE.

A glimpse into the project’s strategic planning, showing a screen outlining the ambitious goal of installing multiple supercomputers, including Isambard-AI’s phases, within 12 months.

A glimpse into the project’s strategic planning, showing a screen outlining the ambitious goal of installing multiple supercomputers, including Isambard-AI’s phases, within 12 months.

A Turning Point for UK Science

Named after Isambard Kingdom Brunel — the 19th-century engineer who reshaped Britain with railways, bridges and ships — Isambard-AI brings that same scale of ambition to AI research.

And like Brunel’s projects during his era, it moves fast. Having gone from conception to deployment in just under two years, including the construction of the modular data center in 48 hours, the supercomputer is already running live projects aligned with national priorities.

Early flagship projects include:

  • Nightingale AI: A sovereign, multimodal health foundation model trained on National Health Service (NHS) data to support earlier diagnoses and personalized care.
  • BritLLM: A U.K.-developed LLM project supporting British languages like Welsh, alongside English, to promote inclusivity and better public service delivery in healthcare, education and public services.
  • UCL Cancer Screening AI: Developing the first scalable AI system for prostate cancer detection via MRI, aiming for faster diagnoses and tailored treatments.
  • EIMCRYSTAL (University of Liverpool): Using AI to search 68 million chemical combinations to discover greener, more sustainable industrial materials — reducing reliance on rare or toxic inputs.
  • EgoAI (University of Bristol): Using AI to analyze recordings from wearable cameras and other smart devices to help people perform tasks better at home. This holds immense promise for assisting dementia patients in the future.
A technician carefully installs a server blade equipped with direct liquid cooling, demonstrating the hands-on precision involved in assembling Isambard-AI’s advanced computing infrastructure.


A technician carefully installs a server blade equipped with direct liquid cooling, demonstrating the hands-on precision involved in assembling Isambard-AI’s advanced computing infrastructure.

Built Fast, Built Smart

How do you build a supercomputer in half the usual time?

“We treated the project like a high-performance processor,” said Simon McIntosh-Smith, director at BriCS. “We executed everything in parallel.”

Key to this speed:

  • Parallelized project management
  • Early procurement of critical hardware, particularly GPUs
  • Prefabricated modular data halls shipped in and assembled in days
  • Relentless optimization of delivery timelines
  • Ground broke in June 2024. By June 2025, the full 5-megawatt system was live: built, tested and running at a global scale.
A close-up of the intricate network of liquid-cooling pipes, a crucial component of Isambard-AI’s world-class energy efficiency and ability to pack 440 GPUs per cabinet.


A close-up of the intricate network of liquid-cooling pipes, a crucial component of Isambard-AI’s world-class energy efficiency and ability to pack 440 GPUs per cabinet.

World-Class Efficiency

Isambard-AI is not just fast — it’s green.

  • Ranked fourth globally on the Green500 list
  • Powered by NVIDIA GH200 Superchips, among the most efficient chips available
  • Direct liquid-cooling architecture from HPE packs 440 GPUs per cabinet
  • Hybrid cooling towers enable >90% dry, waterless operation
  • Entirely powered by zero-carbon electricity
  • Engineered with the potential to recycle waste heat
  • Overall power usage effectiveness below 1.1, among the best ever achieved
An aerial view of the Isambard-AI construction site, illustrating the scale of the modular data center build and the rapid progress made in creating the U.K.’s most powerful AI supercomputer.


An aerial view of the Isambard-AI construction site, illustrating the scale of the modular data center build and the rapid progress made in creating the U.K.’s most powerful AI supercomputer.

A National Asset

Access will be managed by the U.K. Department for Science, Innovation, and Technology and U.K. Research and Innovation, with an emphasis on aligning usage with national priorities, enabling smaller institutions and startups, and supporting both research and commercial innovation.

The U.K. now has an AI infrastructure asset that matches its ambitions, not just for headline moments but for deep, lasting national impact.

The UK’s AI Playbook: Infrastructure, Investment, Acceleration

The launch of Isambard-AI marks a big leap for the U.K., one part of a broader, coordinated push to put Britain at the forefront of AI.

From supercomputers to skills programs, Britain is building the tools to lead in AI — on its own terms.

Images and video courtesy of the Bristol Centre for Supercomputing

Related News



Source link

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Business

AI company Anthropic to pay authors $1.5 billion in landmark settlement

Published

on


Big numbers often get thrown around in the aftermath of legal battles, as judges hand down judgements—or attorneys arrange settlement amounts—in the tens, or hundreds, of millions of dollars. Still, even jaded legal observers can occasionally run into a genuinely daunting number while parsing this stuff. Like, say, the $1.5 billion settlement that AI company Anthropic has agreed to pay in the ongoing class-action suit against it, launched by authors who said the company infringed on their copyrighted works by feeding them as training data to its “AI assistant” Claude. Sure, parts of that sum (calculated at $3,000 per work for a staggering number of works, and with its first $300 million installment due just five days after the settlement is approved) might potentially vanish in a puff of future bankruptcy. But it’s still the “largest publicly reported copyright recovery in history,” according to legal documents from the authors’ attorneys.

That being said, the win here on the wider AI front is quite a bit less clear than “hand our clients the annual estimated GDP of Grenada” might suggest. Yes, U.S. District Judge William Alsup set the stage for Anthropic to eat that massive price tag by ruling that the company clearly violated copyright agreements via how it acquired the books it fed into its own personal woodchipper. (I.e., downloading pirated datasets of millions of books that had been floating around the internet.) And, yes, the settlement will require Anthropic to destroy those “shadow library” datasets in its possession. (But notably, with no actual changes to the Claude large language model itself.) Most critically, though, back in June, Alsup also ruled that “reproducing purchased-and-scanned books to train AI” falls under fair use, calling the case “exceedingly transformative” as a justification for the designation.

As such, both sides in the fight issued statements claiming a form of victory today, with the authors’ side focusing mostly on the massive size of the settlement amount. Anthropic, meanwhile—which has been backed in the past with more than $6 billion in contributions from Amazon and Google—focused its statements on the legal precedent it achieved in the case: “In June, the District Court issued a landmark ruling on AI development and copyright law, finding that Anthropic’s approach to training AI models constitutes fair use. Today’s settlement, if approved, will resolve the plaintiffs’ remaining legacy claims.” What this likely means is that AI companies aren’t going to slow down—especially with, say, a $1.5 billion mortgage suddenly hanging over their heads—but simply become a lot more choosy about how they get their training data.

[via Deadline]




Source link

Continue Reading

Business

Broadcom Inc. Reports Record Revenue Amid AI Growth

Published

on


Broadcom Inc. ((AVGO)) has held its Q3 earnings call. Read on for the main highlights of the call.

Elevate Your Investing Strategy:

  • Take advantage of TipRanks Premium at 50% off! Unlock powerful investing tools, advanced data, and expert analyst insights to help you invest with confidence.

The recent earnings call from Broadcom Inc. showcased a strong performance in AI semiconductors and infrastructure software, with record revenues and a solid backlog. Despite some challenges in the non-AI semiconductor segment and pressures on gross margins due to product mix, the overall sentiment was optimistic. The positive highlights significantly outweighed the lowlights, indicating a promising outlook for future growth, particularly in AI.

Record-Breaking Revenue and Growth

Broadcom Inc. reported a record total revenue of $16 billion, marking a 22% increase year-on-year. This impressive growth was primarily driven by the strong performance in AI semiconductors and the expansion of VMware. The company’s ability to achieve such significant revenue growth underscores its strategic focus on high-growth areas.

AI Semiconductor Growth

The AI semiconductor segment was a standout performer, generating $5.2 billion in revenue, which represents a 63% increase year-on-year. This marks the 10th consecutive quarter of robust growth in this segment. Looking ahead, Broadcom forecasts AI semiconductor revenue to reach approximately $6.2 billion in Q4, up 66% year-on-year, highlighting the company’s leadership in this rapidly expanding market.

Infrastructure Software Segment Performance

Broadcom’s infrastructure software segment also delivered strong results, with revenue reaching $6.8 billion, up 17% year-on-year. The total contract value booked during Q3 was $8.4 billion, reflecting the company’s strength in securing long-term commitments from customers.

Strong Backlog and Bookings

The company’s consolidated backlog reached a record $110 billion, with bookings showing robust growth, particularly in AI. This substantial backlog provides a solid foundation for future revenue and demonstrates strong customer demand across Broadcom’s product lines.

CEO Tenure Extension

In a significant leadership development, Broadcom’s board and CEO Hock Tan have agreed that he will continue as the CEO through at least 2030. This extension provides stability and continuity in leadership, which is crucial for executing the company’s long-term strategic vision.

Non-AI Semiconductor Demand

While the AI segment thrived, the non-AI semiconductor demand remained sluggish, with Q3 revenue of $4 billion flat sequentially. Enterprise networking and service storage experienced sequential declines, with only broadband showing strong growth. This highlights the challenges Broadcom faces in certain segments of its semiconductor business.

Gross Margin Impact

Broadcom anticipates a slight decline in its Q4 consolidated gross margin, down approximately 70 basis points sequentially. This is primarily due to a higher mix of XPUs and wireless revenue, which impacts the overall product mix and margin structure.

Forward-Looking Guidance

During the earnings call, Broadcom provided robust guidance for the upcoming quarter and fiscal year. The company forecasts Q4 2025 consolidated revenue of $17.4 billion, up 24% year-on-year, with AI semiconductor revenue expected to reach $6.2 billion, up 66% year-on-year. Infrastructure software revenue is projected at $6.7 billion, up 15% year-on-year. Broadcom anticipates an adjusted EBITDA margin of 67% for Q4, with continued growth in the AI business and the addition of a significant fourth customer expected to positively impact fiscal 2026.

In summary, Broadcom Inc.’s latest earnings call highlighted a strong performance in AI semiconductors and infrastructure software, with record revenues and a promising outlook for future growth. Despite some challenges in non-AI segments and margin pressures, the overall sentiment was optimistic, driven by significant achievements and robust forward-looking guidance.

Disclaimer & DisclosureReport an Issue



Source link

Continue Reading

Business

Runway founder Cristóbal Valenzuela wants Hollywood to embrace AI

Published

on


At 84, veteran mogul John Malone is still a power broker, hinting at “further consolidation in the media industry” following a recent sit down with David Ellison. Should we be on the lookout for a Warner–Paramount merger? Meanwhile in Vegas, the Sphere’s $100 million Wizard of Oz reimagining leans on AI to expand the visuals and even slip in cameos of David Zaslav and James Dolan. The Directors Guild did not take kindly to the stunt. Partners in Banter Kim Masters and Matt Belloni pull back the curtain on the Sphere’s Emerald City sideshow.

Plus, Masters speaks with Runway co-founder Cristóbal Valenzuela about the role of artificial intelligence in Hollywood. The Chilean-born developer acknowledges that AI may lead to some job losses, but he argues it will ultimately benefit filmmakers. He explains why studios including Lionsgate, Netflix, and Disney are already using Runway’s tools. Plus, he compares the current backlash against AI to the upheaval that followed the introduction of sound in film.





Source link

Continue Reading

Trending