PhysicsX is a London-headquartered company building artificial intelligence models that stand in for numerical physics simulation in engineering design. Its founders come from Formula One rather than software: Robin Tuluie previously led research and development at the Renault and Mercedes F1 teams, and Jacomo Corbo was chief race strategist at Renault and a co-founder of QuantumBlack, the analytics firm acquired by McKinsey. The company has raised capital at a pace unusual for UK industrial software.
A $135m Series B led by Atomico closed in June 2025, taking total funding to nearly $170m; a year later, on 8 June 2026, it announced a $300m Series C led by Temasek at a valuation of roughly $2.4bn. The investor list is notable for who is on it: Siemens, NVIDIA and Applied Materials sit alongside financial backers including M&G Investments, Atomico and General Catalyst. The company says headcount has doubled in the past year to more than 300, with offices in London and New York and a growing presence in the Bay Area and Singapore.
Conventional engineering design relies on computational fluid dynamics and finite element analysis — simulations that can take hours or days per run, limiting how many design variants an engineer can test. PhysicsX trains machine-learning surrogate models on large volumes of simulation data so that performance predictions arrive in seconds; the "hours to milliseconds" framing is the company's, and its accuracy-parity claims have not been independently benchmarked in public. The most concrete evidence of the approach is LGM-Aero, a "large geometry model" for aerospace unveiled in December 2024 at AWS re:Invent.
It was trained on more than 25 million meshes representing over 10 billion vertices, plus tens of thousands of CFD and FEA simulations generated using Siemens Digital Industries Software tools. A free showcase application, Ai.rplane, generates novel aircraft geometries and predicts lift, drag, stability and structural stress, in under a second per design, according to the company. The Siemens relationship goes beyond training data: the two firms have a stated collaboration on AI-based deep physics simulation, and Siemens has invested in successive rounds. Strategic investment from Siemens, NVIDIA and Applied Materials is a meaningful signal of industrial validation, though it is not the same as neutral third-party benchmarking.
PhysicsX is one of the more substantial AI companies actually headquartered in London, and its target sectors map directly onto UK infrastructure priorities. The company lists energy, data centres, aerospace and defence, semiconductors, automotive and industrial machinery among the industries it serves. Two of those matter particularly for this publication's readership: the design of energy hardware, from turbomachinery and heat exchangers to the components that grid and generation projects depend on, and thermal design for data centres, where the UK's build-out is constrained by both power and cooling.
Faster design iteration is not a marginal gain in these sectors; it changes how quickly hardware for the energy transition can be engineered and certified. The presence of M&G, a large UK institutional investor, in the Series C also ties a UK savings institution to the company's growth. What is not yet on the record is a named UK infrastructure customer: the verified relationships are with global industrial partners rather than domestic asset owners.
The company's relevance to UK infrastructure rests today on its headquarters, its sector focus and the Siemens collaboration rather than on disclosed UK deployments — so the confirming signal would be a named customer in UK energy, grid or data-centre engineering, or published case studies with independently checkable results. Third-party benchmarks of surrogate-model accuracy against full simulation would settle the central technical claim.
Points of caution: growth metrics (revenue doubling, tripled bookings) are company-supplied and unaudited; the $300m round implies substantial spending commitments; and expansion into New York, the Bay Area and Singapore raises the familiar question of whether the engineering centre of gravity stays in London. None of these undermines the thesis yet. A UK-headquartered firm at a $2.4bn valuation selling physics AI into energy and data-centre design is worth following on the evidence already in hand.
Sources: markets.financialcontent.com · physicsx.ai · physicsx.ai
Entries are verified against primary sources. The bar is the frontier: first-of-kind, new technology, new materials — delivery milestones of conventional infrastructure, however large, do not qualify. The frontier is tracked globally; every entry states what it means for UK and Irish infrastructure.
Verified by GridteamAI.
