Level 1 - Absolute Beginner

PhysicsX is a company in London, England. It uses AI to help engineers build better machines. Engineers design things like planes, cars, and rocket engines.

When engineers build something, they need to test their designs. Normal tests take many days on powerful computers. PhysicsX uses AI to do the same tests in just seconds.

On June 8, 2026, PhysicsX raised 300 million dollars from investors. Big companies like NVIDIA and Siemens gave money to help PhysicsX grow.

The company was started by people who worked in Formula 1 car racing. It is now worth about 2.4 billion dollars.

engineer

a person who designs and builds machines, buildings, or other things

simulation

a computer test that acts like a real situation to see what will happen

investor

a person or company that gives money to a business hoping to earn more money back

AI (artificial intelligence)

computer programs that can learn and solve problems like a human

design

a plan that shows how something will look or work when it is built

raise

to collect or obtain money for a specific purpose

worth

how much money something is valued at

stress

a force that pushes, pulls, or bends an object and can damage it

Level 2 - Elementary

PhysicsX, a British AI company based in London, raised $300 million in a Series C funding round on June 8, 2026. The company makes AI models that can simulate how materials and parts behave when exposed to heat, pressure, and airflow. These simulations traditionally take days on powerful computers, but PhysicsX can produce results in seconds.

The company calls its technology Large Physics Models (LPMs). These are similar to the large language models that power AI chatbots like ChatGPT, but instead of being trained on words and text, they are trained on physics data. This allows them to predict how a jet engine, car part, or turbine will perform under real conditions.

The funding round was led by Temasek, Singapore's government investment fund. Other investors included NVIDIA, Applied Materials, Siemens, and several venture capital firms. The round valued PhysicsX at approximately $2.4 billion, more than double its valuation from a year earlier.

PhysicsX was founded in 2019 by Jacomo Corbo and Robin Tuluie, both former Formula 1 engineers. Its customers include companies in the aerospace, automotive, and nuclear energy industries. The company doubled its revenue and tripled its booked orders in the past year.

Series C

the third major round of investment funding for a growing company

Large Physics Model (LPM)

an AI model trained on physics data to simulate how objects behave in the real world

venture capital

money invested by specialist firms in new companies with high growth potential

valuation

the estimated total monetary value of a company

aerospace

the industry relating to the design and construction of aircraft and spacecraft

turbine

a machine with spinning blades driven by air, water, or steam to produce power

nuclear energy

electricity produced from the heat generated by splitting atoms inside a reactor

booked orders

contracts signed by customers that guarantee future revenue for a company

Level 3 - Intermediate

PhysicsX, the London-based AI company building what it calls Large Physics Models (LPMs), secured a $300 million oversubscribed Series C on June 8, 2026, placing its valuation at approximately $2.4 billion. The round was led by Temasek, Singapore's sovereign wealth fund, with existing backers NVIDIA, Applied Materials, Atomico, General Catalyst, and Siemens all increasing their stakes, reflecting broad confidence in AI-driven engineering simulation.

LPMs are pre-trained foundation models analogous to large language models but built on physics data rather than text. The core proposition is that an LPM can evaluate how a component behaves under stress, thermal load, or aerodynamic forces in seconds rather than the hours or days required by conventional computational fluid dynamics or finite-element analysis software. This compression of simulation time translates directly into faster design iteration and lower prototyping costs for customers in aerospace, automotive, and energy.

PhysicsX was founded in 2019 by Jacomo Corbo, a former co-founder of McKinsey's QuantumBlack AI division, and Robin Tuluie, former head of R&D at Renault Alpine F1. The Formula 1 origin is significant: the sport's engineering culture demands extremely fast iterative simulation to optimize aerodynamic performance within strict regulatory constraints, creating precisely the use case that LPMs are designed to address.

The company reported doubling year-on-year recognized revenue, tripling booked revenue, and more than doubling its customer count in the twelve months before the Series C. With the fresh capital, PhysicsX plans to accelerate frontier research into increasingly powerful LPMs and expand into additional multi-physics domains, including structural mechanics, electromagnetics, and thermal management.

sovereign wealth fund

a state-owned investment fund that manages a country's financial reserves and national savings

foundation model

a large AI model trained on broad data that can be adapted for many specific applications

computational fluid dynamics (CFD)

a branch of engineering that uses numerical calculations to simulate how fluids move and interact

finite-element analysis

a computational method that divides a structure into small segments to predict how it responds to forces

iteration

one complete cycle of a repeated process, especially when each cycle improves on the last

thermal load

the amount of heat energy that a material or structure must absorb and withstand

aerodynamic

relating to how air flows around an object and the forces this creates on it

regulatory constraints

limits and rules imposed by official governing bodies that must be followed

Level 4 - Advanced

PhysicsX's $300 million oversubscribed Series C, led by Temasek with co-participations from NVIDIA, Applied Materials, Atomico, General Catalyst, and Siemens, marks the most consequential funding event yet for the nascent Large Physics Model (LPM) category. At a $2.4 billion post-money valuation, the company has more than doubled its worth from sub-$1 billion twelve months prior, underscoring investor conviction that AI-accelerated engineering simulation represents a structural displacement of conventional solver software across capital-intensive industries.

The LPM paradigm shares its architectural grammar with the transformer-based foundation models that dominate natural-language and vision AI, but substitutes physics-domain training corpora for text and image datasets. The consequence is a model capable of evaluating stress, thermal, and aerodynamic responses across a component's parameter space in wall-clock times measured in seconds, compressing the iteration cycle that traditionally required hours of finite-element analysis or computational fluid dynamics on dedicated high-performance clusters. For sectors such as aerospace, where a single turbine-blade design iteration can cost tens of thousands of dollars in compute time, the margin economics are transformative.

The founders' Formula 1 provenance is not merely biographical color. F1 engineering operates at the intersection of extreme simulation density and narrow regulatory windows: teams run hundreds of aerodynamic iterations per design cycle under CFD restrictions imposed by the FIA, making sub-second LPM inference a genuine competitive differentiator rather than a convenience feature. Corbo's prior role leading QuantumBlack, McKinsey's AI unit, provided the organizational and commercial playbook for scaling a deep-tech product into enterprise workflows at speed.

The company's tripling of booked revenue and doubling of recognized revenue in the fiscal year preceding the Series C signals that adoption has moved beyond pilot engagements into multi-year enterprise contracts. With new capital earmarked for frontier LPM research across multi-physics domains, global headcount expansion, and platform integrations with major CAD and PLM ecosystems, PhysicsX is positioning itself as the inference layer for industrial engineering in a manner directly analogous to how OpenAI positioned itself as the inference layer for enterprise knowledge work.

nascent

just beginning to exist or develop; not yet fully established or mature

paradigm

a fundamental model or framework that defines the standard approach in a field

architectural grammar

the foundational structural principles that define how an AI model or technical system is built

wall-clock time

the actual real-world elapsed time a computation takes, as measured on a clock

provenance

the origin or source of something; the history of where and how it came from