PhysicsX的赌注：AI预测物理速度快至100万倍；3200万美元A轮融资，三年估值达24亿美元

做的是什么生意

AI-native engineering software that runs physics simulations in seconds, sold to aerospace, automotive, semiconductor, energy and materials manufacturers.

启动资金：$32M Series A (Nov 2023), first outside funding, led by General Catalyst with Standard Investments, NGP, Radius Capital and Henry Kravis

起因

PhysicsX was founded in 2023 in London by Robin Tuluie, former head of R&D at Renault (Alpine) F1 and chief scientist at Mercedes F1, and Jacomo Corbo, founder of McKinsey's QuantumBlack and a former F1 chief race strategist. Their thesis: traditional simulation (CFD/FEA) takes hours to days per design, capping how many variants engineers can test. The company emerged from stealth on 27 Nov 2023 with a $32M Series A led by General Catalyst and a team of about 50 scientists, claiming its AI models predict physical behavior 10,000 to a million times faster than conventional simulation software.

经过

Between 2023 and 2025 PhysicsX embedded its platform with aerospace, defense, automotive, semiconductor, energy and materials customers. In June 2025 it raised a $135M Series B led by Atomico with Temasek, Siemens, Applied Materials and July Fund, bringing total funding to nearly $170M; the team grew past 150 and revenue more than quadrupled in two years. In June 2026 it raised an oversubscribed $300M Series C led by Temasek — joined by M&G Investments and Intrepid Growth Partners and existing backers including NVIDIA and Siemens — at a valuation of roughly $2.4B, with revenue doubled year over year, booked revenue tripled and customer count more than doubled.

结果

Still scaling: by June 2026 PhysicsX had more than 300 employees, offices in London, New York, the Bay Area and Singapore, and was spending the new capital on 'Large Physics Models' — larger pre-trained physics AI models — while expanding across industrial sectors.

可借鉴之处

对现有工具进行数量级加速可以创造新类别：攻击关键工作流中最慢的步骤，从创始人有内部信誉的地方开始。