This PhysicsX company profile covers the company's valuation, revenue, funding rounds, investors and history. PhysicsX works in AI, with a focus on engineering simulation. It was founded in 2019 and is headquartered in London, United Kingdom. Its latest valuation on record is US$2.4 billion (2026). Every figure links to a public source, and anything the company hasn't disclosed is marked as such.
Revenue
Funding history
| Date | Round | Amount | Valuation | Investors | Source |
|---|---|---|---|---|---|
| 2026 | Series C | US$300M | US$2.4B | Temasek (lead) | Sifted |
| Jun 2025 | Series B | US$135M | Close to US$1B | Atomico (lead); later extension with NVentures | The Logic |
PhysicsX timeline
Key dates from the sources on this page, oldest first. Metric snapshots are listed after the funding events.
- 2019 — PhysicsX is founded in London by Robin Tuluie and Jacomo Corbo (CEO).
- Jun 2025 — Series B: US$135M, valuing the company at Close to US$1B. Investors: Atomico (lead); later extension with NVentures.
- 2026 — Series C: US$300M, valuing the company at US$2.4B. Investors: Temasek (lead).
Investors
Overview
PhysicsX builds AI that helps engineers design physical products, such as aircraft parts, car components, semiconductors and energy equipment, much faster. Engineers traditionally rely on numerical simulations, such as computational fluid dynamics and finite element analysis, to test designs virtually. These simulations can take hours or days to run, which limits how many design options engineers can explore. PhysicsX trains AI models on simulation and test data so they can predict how a design will perform in seconds, letting engineers evaluate thousands of options and optimise designs in ways that were previously impractical.
PhysicsX was founded in London in 2019 by Robin Tuluie, a physicist who previously led research and development at Formula One teams, and Jacomo Corbo, formerly of McKinsey's QuantumBlack AI unit. It raised a US$135 million Series B led by Atomico in 2025 at a valuation close to US$1 billion, later extended with Nvidia's venture arm NVentures. In 2026 it raised US$300 million led by Singapore's Temasek, more than doubling its valuation to US$2.4 billion, according to Sifted.
Products
- Large physics models: AI models trained on physics simulations that generalise across designs.
- Engineering platform: software for engineers to build, run and optimise AI-accelerated simulations.
- Design optimisation: tools that search large design spaces for the best-performing options.
- Industry solutions: applications for aerospace, automotive, semiconductors, energy and defence.
Funding
PhysicsX raised a Series A in 2023 with investors including General Catalyst. In June 2025 it raised a US$135 million Series B led by Atomico, which reports said valued it close to US$1 billion. Later in 2025 NVentures joined in an extension. In 2026 it raised a US$300 million round led by Temasek at a US$2.4 billion valuation, as reported by Sifted and Trending Topics.
Business model
PhysicsX sells software subscriptions and enterprise contracts to engineering-intensive companies, often beginning with projects that demonstrate value on specific design problems and then expanding to platform deployments. Customers pay for software access, model development and support.
Competition
PhysicsX competes with traditional simulation software vendors such as Ansys (now part of Synopsys), Siemens, Dassault Systèmes and Altair, which are adding AI features; with AI simulation startups such as Neural Concept and Navier AI; and with customers' internal teams. Its differentiation is a focus on AI-native engineering at scale.
PhysicsX at a glance
PhysicsX is a London physical AI company founded in 2019 by Robin Tuluie and Jacomo Corbo. It builds AI models that predict how engineered products will perform, accelerating design in aerospace, automotive, semiconductors and energy. It raised a US$135 million Series B led by Atomico in 2025 and a US$300 million round led by Temasek at a US$2.4 billion valuation in 2026.
Timeline
- 2019: founded in London.
- 2023: raises a Series A.
- June 2025: raises a US$135 million Series B led by Atomico.
- Late 2025: NVentures joins in an extension.
- 2026: raises US$300 million led by Temasek at US$2.4 billion.
Why AI simulation matters
Designing a jet engine part or an electric vehicle battery involves balancing many factors: weight, strength, heat, airflow, cost and manufacturability. Each design change requires new simulations. Because simulations are slow, engineers typically test only a handful of options. AI models trained on simulation data can approximate results almost instantly, enabling engineers to explore thousands of variants and find designs humans might not have considered. This can shorten development cycles and improve performance. The challenge is ensuring AI predictions are accurate enough to trust, particularly for safety-critical products.
Formula One roots
Formula One teams operate under strict limits on wind tunnel time and computing, so they have long pushed simulation and optimisation to the limit to gain small performance advantages. Tuluie's experience in motorsport engineering shaped PhysicsX's approach: combining physics expertise with data-driven methods to iterate quickly. The company has hired many engineers from motorsport, aerospace and scientific computing.
Who funds PhysicsX
PhysicsX's investors include Temasek, Atomico, NVentures and General Catalyst.
Valuation context
Our guides to the largest UK startups and the most valuable AI startups compare PhysicsX with peers.
Risks
- Accuracy and trust: engineers need confidence in AI predictions.
- Incumbent competition: simulation giants are adding AI.
- Long sales cycles: industrial customers adopt new tools slowly.
- Data access: models depend on customers' proprietary data.
Questions for engineering teams
- Which design problems take the most simulation time today?
- How will AI predictions be validated?
- Can models be trained on your proprietary data securely?
- How does the platform integrate with existing CAD and simulation tools?
An aerospace example
An aircraft engine maker wants to redesign a component to reduce weight while handling high temperatures. Using traditional simulation, each design iteration might take a day. With PhysicsX, engineers train a model on existing simulation results, then use it to evaluate thousands of design variants in hours. The best candidates are verified with full simulations and physical tests. The team arrives at a lighter design in weeks rather than months.
Physical AI as a category
Much of the attention in AI has focused on language and images, but a growing group of companies applies AI to the physical world: robots, materials, chemistry, and engineering design. PhysicsX belongs to this "physical AI" category. The economic stakes are large: industries such as aerospace, automotive, energy and semiconductors spend heavily on research, development and testing, and shortening design cycles can bring products to market faster and improve performance. Unlike consumer AI, these customers demand rigorous validation, and mistakes can be costly or dangerous. Companies that earn trust in these industries can build durable relationships.
The category also benefits from advances in computing hardware. Nvidia has promoted AI-accelerated simulation and digital twins, and its venture arm's investment in PhysicsX reflects interest in expanding AI beyond data centres into industrial design. Collaboration with hardware providers helps PhysicsX run large models efficiently.
Defence and strategic industries
Reports on PhysicsX's funding noted interest from defence-related sectors, where faster design of aircraft, drones, propulsion systems and other equipment is a priority amid rising military spending in Europe. Engineering AI could help defence manufacturers develop new systems more quickly. This creates opportunities but also brings scrutiny over export controls and ethical considerations. Our guide to defence tech startups covers related companies.
What to watch
Watch for large industrial customer deployments, defence applications, new funding and how incumbents respond.
Takeaways
Engineering teams can explore far more designs with AI. Founders can see how domain expertise and AI combine. Investors should watch validation and adoption in industry.
Sources
Every figure on this page comes from the public sources below. Where sources disagreed, we used the company's own announcement or a primary news report. Private-company data changes; check the date beside each figure.
Frequently asked questions
What does PhysicsX do?
PhysicsX is a ai / engineering simulation company headquartered in London, United Kingdom. It was founded in 2019.
What is PhysicsX's valuation?
PhysicsX was valued at US$2.4 billion as of 2026, according to Sifted. Private company valuations are set at funding rounds or share sales and can change significantly between them.
How much revenue does PhysicsX make?
PhysicsX is a private company and has not publicly disclosed its revenue. Figures on third-party sites are usually modelled estimates rather than numbers reported by the company.
What was PhysicsX's latest funding round?
The most recent financing we have verified is a series c in 2026: US$300M, at a valuation of US$2.4B. Source: Sifted.
Who are PhysicsX's investors?
Investors named in public announcements include Temasek, Atomico, NVentures (Nvidia) and General Catalyst.
Who founded PhysicsX?
PhysicsX was founded in 2019 by Robin Tuluie and Jacomo Corbo (CEO).
Is PhysicsX a public company?
No. PhysicsX is a private company (private, venture backed), so its shares are not traded on a stock exchange and it isn't required to publish its financial results.
Where is PhysicsX headquartered?
PhysicsX is headquartered in London, United Kingdom.