Tag: CUDA

NVIDIA Open Sources First GPU-Accelerated Medical Physics Simulation Framework

Posted on by faz_business



Before a healthcare robot can be useful in the real world, it has to learn how the physical world pushes back. Anatomy varies. Instruments bend, press, slip and interact with tissue. Imaging can be noisy or incomplete. And the rare, edge scenarios developers most need to understand don’t appear on schedule.

That creates one of the biggest bottlenecks in healthcare robotics: obtaining the enormous amount of varied data developers need to train, test and improve robot behavior.  

NVIDIA Medical Physics Simulation framework — a new open source, GPU-accelerated capability within NVIDIA Isaac for Healthcare — announced today, helps medical robotics developers model anatomy-device interaction, generate hard-to-capture scenarios, test in silico, and train or evaluate robot policies before hardware-heavy testing. 

The framework brings together anatomy and medical device behavior with sensor simulation and robot learning so teams can create reusable simulation environments instead of rebuilding custom scenes for every workflow, saving developers time and bringing innovations to market faster. 

Because Medical Physics Simulation is open source, healthcare robotics developers can inspect the framework, adapt it to their own devices and workflows, and build on a GPU-accelerated foundation that works seamlessly with the broader NVIDIA stack.

Open source is especially important in healthcare because teams need transparency into the data, models and weights that shape system behavior. Access to open models and model weights can help developers reproduce results, evaluate performance across different anatomies and scenarios, identify limitations and build evidence for regulatory review. 

A Virtual Training Ground for Medical Robots

For physical AI, experience is data in motion. Developers need to train robots to operate properly even when anatomy changes, devices behave differently, conditions shift or a policy fails unexpectedly.

Medical Physics Simulation helps developers simulate anatomy, device contact, friction and sensor inputs, then test in interactions and environments to evaluate how robots perform across those changes. Powered by NVIDIA CUDA and part of Isaac for Healthcare — built on the NVIDIA Warp, Newton and Cosmos simulation and generative AI technologies — the framework can run hundreds of parallel simulation environments, helping teams explore more scenarios and identify failure modes earlier in development. 

For robot builders, this turns simulation from a bespoke engineering project into reusable infrastructure. The difference now is scale: benchmarks show 8,192 robot-training environments running in parallel with GPU-native simulation cut training from over five hours to under two minutes. 

With this framework, developers can connect vascular anatomy, flexible instruments such as catheters and guidewires, simulated X-ray imaging and reinforcement learning. The framework is designed to extend beyond that example to additional devices, anatomies, sensors and healthcare robotics domains.

Medical Physics Simulation brings together classical physics simulation and generative AI physics simulation. Classical simulation helps model known physical rules, such as device contact, friction and motion. NVIDIA Cosmos-H Dreams, the real-time generative AI physics simulation capability within Medical Physics Simulation, helps model visual scene dynamics learned from procedural data.

Together, these approaches give developers a richer way to build and test healthcare robotics systems in virtual environments before moving to physical prototypes and lab testing.

An Ecosystem Building the Future of Medical Robotics

Medical robotics leaders are already applying simulation-driven development to solve specific surgical challenges.

CMR Surgical and Cambridge Consultants, part of Capgemini, are using Cosmos-H-Dreams to implicitly learn interaction physics for soft-tissue surgical procedures and generate patient-specific simulations. CMR contributed nearly 500 hours of anonymized clinical data from its Versius Surgical Robotic System to the Open-H Embodiment open dataset, benefiting procedures including cholecystectomy, prostatectomy, hernia repair and hysterectomy.

“Open source models allow us to build on shared knowledge, accelerating responsible innovation and, ultimately, gives us the potential to deliver more consistent care and better outcomes for patients worldwide,” said Chris Fryer, chief technology officer at CMR Surgical.

Johnson & Johnson MedTech is using Isaac for Healthcare’s Medical Physics Simulation and a Cosmos-based foundation model to build digital twins of its endoluminal MONARCH platform for urology, modeling complex anatomy and kidney-stone scenarios.

XCath is using the Medical Physics Simulation for endovascular autonomy policy training. Inner Logic is accelerating the evolution of medical technology with synthetic data, validating device mechanics and producing in silico evidence to support regulatory pathways with NVIDIA Medical Physical Simulation.

Medtronic Structural Heart is exploring applying Medical Physics Simulation with simulated X-ray sensing to generate data for catheter navigation research.

A New Layer in the Isaac for Healthcare Stack

As a modular capability within NVIDIA Isaac for Healthcare, Medical Physics Simulation can be used on its own or alongside digital twin pipelines, medical sensor simulation, the NVIDIA Isaac Lab open robot-learning framework and NVIDIA open models and policies.

Developers can explore the open source Medical Physics Simulation framework, review available reference workflows and start building simulation environments for their own devices, anatomies and healthcare robotics applications.

NVIDIA Blackwell Leads on First Agentic AI Infrastructure Benchmark

Posted on by faz_business


AgentPerf from Artificial Analysis, the industry’s first agentic AI benchmark, gives developers, enterprises and infrastructure providers a clear way to compare systems for agentic AI. In the first round of published results, the NVIDIA Blackwell Ultra NVL72 platform delivers leading performance across the agentic AI workloads tested, running 20x more agents per megawatt than NVIDIA Hopper.

Agentic AI is a fundamentally different workload than conversational AI. A single chat completion is a sprint: one large language model (LLM) call, one response. An agent functions more like a relay: It breaks a goal into many steps and keeps going until the task is done. 

Agents chain together multiple LLM calls and tool calls to gather context, observe, reason and act.

That results in dozens to hundreds of LLM calls chained together, each passing growing context to the next, with tool calls like code compile and execution, database search and web browsing at every handoff. The complexity isn’t additive; it’s multiplicative. 

The distinction matters enormously for performance measurement. Existing AI inference benchmarks measure one LLM call: how fast an LLM responds to a single request and how many simultaneous requests a system can handle. They weren’t designed for agentic workloads, where chained LLM calls, tool call delays and growing context stress accelerated computing systems in fundamentally different ways than a single LLM call ever could. 

For companies building and deploying agents at scale, it’s important to understand how responsive agents are, how many can be deployed simultaneously and how much useful work AI infrastructure can deliver for every dollar and watt invested.

NVIDIA GB300 NVL72 Runs 20x More Agents per Megawatt

In this first round, AgentPerf measures agentic performance with DeepSeek V4 Pro, a large mixture-of-experts (MoE) model that represents the class of frontier models powering today’s most capable agents. On this workload, NVIDIA GB300 NVL72 delivers the highest performance in the benchmark, running up to 20x more agents per megawatt than the NVIDIA HGX H200 system.

NVIDIA GB300 NVL72 supports far more concurrent agents per megawatt than NVIDIA H200 at both service-level objectives of 20 and 60 tokens per second per agent.

The performance advantage comes from extreme codesign across the full stack. GB300 NVL72 connects 72 GPUs into a single rack-scale system, enabling large MoE models like DeepSeek V4 Pro to distribute model execution efficiently at scale. 

CUDA kernels accelerate this further by overlapping communication and compute, so the cost of coordinating across experts is absorbed rather than added to latency. 

NVIDIA TensorRT LLM sustains efficiency as concurrent agent sessions scale. For example, it separates the processing of inputs from the generation of outputs so each can be optimized independently. 

These results are grounded in a benchmark methodology built from the ground up to reflect how agentic AI actually works in production.

Artificial Analysis AgentPerf: Built on Real-World Agentic Workloads

AgentPerf is built based on real coding agent trajectories: an agent receives a task, reads files, writes and edits code, executes commands and iterates based on the results — all drawn from real public code repositories across 12+ programming languages. The long sequence lengths, tool call patterns and delays are all representative of real-world coding workflows. 

AgentPerf then measures how many of these agentic tasks a platform can support simultaneously while meeting defined performance thresholds for responsiveness and output token rate. Tool calls are not executed but simulated using representative CPU processing time, so differences in results reflect accelerated computing performance only. 

The results translate directly into infrastructure decisions: how many concurrent agentic tasks can be run per accelerator and per megawatt of power. For enterprises deploying AI agents at scale, those numbers determine how much productive work a given infrastructure investment can actually deliver.

NVIDIA Ecosystem Partners Harness Blackwell’s Leading Performance

Leading inference providers including Baseten, DeepInfra and Together AI are already serving agentic workloads on frontier models such as DeepSeek V4 Pro on NVIDIA Blackwell and powering production agentic applications today. 

Together AI powers real-time inference for Cursor, an AI-powered agentic coding platform, on NVIDIA Blackwell. Cursor’s agents debug issues, generate features and execute refactors while developers continue working.  

DeepInfra powers Pam.ai, an AI workforce platform for car dealerships, which deploys agents to book service appointments, handle calls and run outbound sales campaigns, entirely on NVIDIA Blackwell. 

As NVIDIA and the open source ecosystem continue to optimize inference software, performance and efficiency on agentic workloads will only improve. The NVIDIA Vera Rubin architecture is now in full production, bringing the next generation of infrastructure capacity to meet the growing demands of agentic AI at scale. 

Dive deeper into AgentPerf’s methodology and NVIDIA’s full-stack optimizations for agentic AI in this technical blog.

How the UK Is Turning Sovereign AI Ambition Into Action With NVIDIA Technologies

Posted on by faz_business


A year ago at London Tech Week, NVIDIA founder and CEO Jensen Huang and U.K. Prime Minister Keir Starmer made a declaration: the U.K. would be an AI maker, not an AI taker. 

At this year’s event, NVIDIA and its partners are showcasing how that commitment is producing real momentum across the nation’s infrastructure, startups and enterprises. 

U.K. technology leaders are innovating across healthcare and life sciences, coding, agentic AI, inference and more — all running on sovereign AI deployments.

AI Minister Kanishka Narayan said: “A year ago, we said the UK would be an AI maker, not an AI taker. Today we’re delivering on that — with sovereign compute powering British startups to push the boundaries of what AI can do, from drug discovery to healthcare to robotics. This is what it looks like when a country backs its own talent with the infrastructure to match.

“NVIDIA’s decision to invest billions here is a reflection of the strength of what’s being built in Britain. We are determined to make sure the next generation of AI breakthroughs happens in this country, and we have everything we need to make it happen.”

Commitment to Compute

Over the past year, the number of AI cloud providers planning to deploy AI infrastructure on U.K. soil has doubled. 

Nebius has announced plans to expand customers and cloud capabilities with three new deployments of advanced NVIDIA AI infrastructure, as the NVIDIA AI Cloud ecosystem partner continues to build out its commercial and AI R&D hub in London. Combined, the deployments are expected to reach 65 megawatts when fully ramped up in 2027.

CoreWeave is building in the U.K. Government’s AI Growth Zones, and seven more NVIDIA AI Cloud ecosystem partners have plans in the pipeline. BT and Nscale announced plans to build sovereign AI data centers across three existing BT sites in the U.K., combining NVIDIA AI infrastructure, Nscale’s full stack and BT’s trusted nationwide connectivity backbone. 

From Fund to Frontier

Central to that sovereign compute story is Isambard-AI — the U.K.’s most powerful computer. Built on 5,400 NVIDIA GH200 Grace Hopper Superchips and running entirely on zero-carbon electricity, it’s the engine behind some of the U.K.’s most ambitious AI research. 

The U.K. government’s Sovereign AI Fund is putting that capability to work by backing homegrown companies and providing the domestic infrastructure needed to scale their ambitions. 

Among its first recipients is Ineffable Intelligence, which recently announced a collaboration with NVIDIA to build the future of reinforcement learning infrastructure. 

Other recipients include four U.K.-based NVIDIA Inception startups, each pushing the AI frontier using Isambard-AI. These startups are:

Cosine Builds Sovereign Coding Platform

Cosine is building an end-to-end sovereign AI coding platform for highly regulated industries such as financial services, critical infrastructure and national security. Using Isambard, Cosine is training a new, large-parameter, mixture-of-experts, multimodal agentic LLM for natively handling data types beyond text and image. 

“Access to Isambard enables the project, full stop,” said Alistair Pullen, cofounder and CEO of Cosine. “We already have the people who know how to do this. We have the data. We have the infrastructure and the training. The thing we’ve never had is this level of compute.”

Cursive Trains Self-Improving AI Systems

Cursive is building self-improving AI systems that learn continuously from real-world data, enabling them to operate autonomously over long periods of time. This is unlocked through new memory-augmented architectures with dramatically larger context windows, currently in development using the Sovereign AI Fund resources. In addition, the team recently adopted the NVIDIA Megatron-LM framework for distributed training at scale.

“The Sovereign AI Fund is more than just processing power — it’s a statement about investing in AI in the U.K.,” said Talfan Evans, cofounder and CEO of Cursive. “Sovereignty is actually now a buying criterion — and it’s a challenge to tap into the resources we uniquely have as U.K. and European companies.”

Doubleword Optimizes Inference to Deliver Abundant Intelligence Tokens

Doubleword, the U.K.’s first dedicated inference lab, optimizes every layer of the AI stack to maximize what it calls “IQ per dollar.” The company deploys open models including NVIDIA Nemotron 3 Super 120B and builds on the NVIDIA Dynamo inference framework. 

On Isambard, Doubleword’s early results achieved 70x faster model cold starts — aka model loading times — and 4x lossless KV cache compression, critical advancements for long-running agentic workloads. The result: inference at 90-95% lower costs than other leading inference providers.

Image courtesy of Doubleword.

“Sovereign AI is most impactful at the inference layer,” said Meryem Arik, cofounder and CEO of Doubleword. “Inference is when you’re actually getting the value from the model — we want that value created in the U.K., with U.K. compute and U.K. data centers.”

Prima Mente Uses Foundation Models to Study Alzheimer’s and More

Prima Mente builds biological foundation models to identify new biomarkers, subtypes and drug targets of Alzheimer’s, Parkinson’s and ALS. With its Isambard allocation, the company is developing Pleiades 2, a foundation model combining five biological data modalities. 

Achieving nearly 3x speedups in model training with NVIDIA Blackwell GPUs, Prima Mente also uses NVIDIA Parabricks for genomic data processing and NVIDIA Transformer Engine for model optimization.

“Research shows Alzheimer’s might be 25 different subgroups of disease, and we want to help by using AI to identify these subtypes and the biology within the cells as they change,” said Hannah Madan, cofounder of Prima Mente.

Video courtesy of Nebius and Prima Mente.

AI Talent, Policy and Production

NVIDIA’s £2 billion investment in the U.K. startup ecosystem — in collaboration with leading venture capital firms — is bringing new capital and advanced AI infrastructure to major U.K. hubs including London, Oxford, Cambridge and Manchester. 

U.K. membership in the NVIDIA Inception program has increased by 50% over the past year. AI-native companies like Doubleword, Synthesia and PolyAI are scaling globally from U.K. roots. 

At last year’s London Tech Week, NVIDIA announced a collaboration with the U.K Department for Science, Innovation and Technology on 6G and AI skills. The 6G collaboration has seeded testbeds at four U.K. universities. In May, the NVIDIA Deep Learning Institute (DLI) delivered two new courses — added to support the nation’s wireless research community — to participants from over 30 U.K. universities.

Plus, as part of this AI skills collaboration, NVIDIA DLI courses are offered as part of QA’s AI Apprenticeships in England. 

And the NVIDIA Developer Program now includes more than 200,000 U.K. developers. 

The Sovereign AI Forum, which launched last year with seven charter members, convened the country’s AI leadership to turn policy into deployment roadmaps. Over the past year, the Forum has welcomed dozens of participants across government, industry and the startup community — turning policy into deployment roadmaps.

And enterprise AI is moving from pilot to production:

  • Apian is building digital twins of two National Health Service hospitals, combining autonomous devices, ground robots, computer vision and robotic simulation.
  • Deliverance AI is helping regulated enterprises to run, govern and scale AI agents inside their own environment — through a single control plane. The Agentic Operating System is built for organizations where data sovereignty is non-negotiable.
  • Glass Futures has installed an AI-driven digital twin of its glass furnace capable of testing and predicting new, optimal ways to make glass. The digital twin taps into NVIDIA accelerated computing and the NVIDIA PhysicsNeMo framework.
  • Orbital Industries has announced codesigned, NVIDIA Vera Rubin DSX AI Factory-compliant AI infrastructure that accelerates time to first token.
  • Reading Football Club is partnering with Stelia to establish an AI Centre of Excellence, combining Stelia’s full-stack AI platform with accelerated compute infrastructure from NVIDIA and Lenovo.

It all reflects momentous progress in U.K. AI leadership — and offers a glimpse of where it’s heading.

Join NVIDIA at London Tech Week.

Everything Will Be Represented in a Virtual Twin, Jensen Huang Says at 3DEXPERIENCE World

Posted on by faz_business



At 3DEXPERIENCE World in Houston, NVIDIA founder and CEO Jensen Huang and Dassault Systèmes CEO Pascal Daloz laid out a blueprint for industrial AI rooted in physics-based “world models” — systems designed to simulate products, factories and even biological systems before they’re built.

“Artificial intelligence will be infrastructure,”  like water, electricity, and the internet Huang told the crowd, playfully referring to the engineering-heavy audience as “Solid Workers,” a nod to Dassault Systèmes’ SolidWorks platform.

The announcement continues a collaboration spanning more than a quarter century between NVIDIA and Dassault Systèmes.

“This is the largest collaboration our two companies have ever had in over a quarter century,” Huang said. “We’re going to fuse these technologies so engineers can work at a scale that’s 100 times, 1,000 times — and eventually a million times greater than before.”

The new partnership brings NVIDIA accelerated computing and AI libraries together with Dassault Systèmes’ Virtual Twin platforms to move more engineering work into real-time digital workflows, powered by AI companions that help teams explore, validate, prototype and iterate faster.

Huang framed the shift as a reinvention of the computing stack: moving from hand-specified, structured digital designs to systems that can generate, simulate and optimize in software — at industrial scale.

From Digital Models to Industry World Models

Virtual twins are not applications, “they are knowledge factories,” Daloz said.

The partnership aims to establish industry world models — science-validated AI systems grounded in physics that can serve as mission-critical platforms across biology, materials science, engineering and manufacturing.

In Daloz’s framing, the value moves upstream: virtual twins become the place where knowledge is created, tested, and trusted — before anything is built in the physical world.

Dassault Systèmes, whose 3DEXPERIENCE platform serves more than 45 million users and 400,000 customers globally, has long been a leader in virtual twin technology — digital replicas that let engineers simulate products and processes before building them physically.

The collaboration brings together accelerated computing, AI and digital twin technologies so engineers can design not only geometry, but behavior — and explore radically larger design spaces earlier in development.

Together, the companies outlined how this shared architecture will show up across science, engineering and manufacturing workflows:

  • Advancing Biology and Materials Research​: The NVIDIA BioNeMo platform and BIOVIA science-validated world models accelerate the discovery of new molecules and next-generation materials.
  • AI-Driven Design and Engineering: SIMULIA AI-based Virtual Twin Physics Behavior leveraging NVIDIA CUDA-X libraries and AI physics libraries empowers designers and engineers to accurately and instantly predict outcomes.
  • Virtual Twins for Every Factory: NVIDIA Omniverse physical AI libraries integrated into the DELMIA Virtual Twin enable autonomous, software-defined production systems.
  • Virtual Companions Supercharge Dassault Systèmes’ Users: The 3DEXPERIENCE agentic platform, combining NVIDIA AI technologies and NVIDIA Nemotron open models with Dassault Systèmes’ Industry World Models, powers Virtual Companions to tap into deep industrial context, delivering trusted, actionable intelligence.

Huang said that in domains like biology and materials, the frontier is learning the underlying “language” of complex systems and then generating new options that can be evaluated and validated in simulation.

Designing and Operating the Factory in Software

A central theme of the discussion was how factories themselves are changing — from static physical assets to living systems that are designed, simulated and operated as virtual twins.

As part of the partnership, Dassault Systèmes is deploying NVIDIA-powered AI factories on three continents through its OUTSCALE sovereign cloud, enabling customers to run AI workloads while maintaining data residency and security requirements.

Both executives emphasized that the goal isn’t to replace engineers — it’s to amplify them. As AI agent companions take on more exploratory and repetitive tasks, designers and engineers gain leverage and creativity, not redundancy.

AI Companions That Expand Human Creativity

Every designer will have a “team of companions,” Huang said — a shift he described as fundamentally positive for engineers, software platforms and the broader ecosystem built on them.

For the tens of millions of engineers who use Dassault Systèmes tools to design everything from aircraft to consumer packaged goods, the shift isn’t about replacing human creativity — it’s about expanding it.

“Success is not about automation,” Daloz said. “[Engineers] don’t want to automate the past — they want to invent the future.”

Looking ahead, Daloz framed the partnership as about more than performance gains – it’s an effort to open new possibilities, help companies eliminate bad choices before they become expensive mistakes, and create entirely new categories of products.

“Virtual twins and the 3D Universes are not applications,” Daloz said. “They are knowledge factories.”

The fireside conversation between Huang and Daloz was broadcast live from 3DEXPERIENCE World.