What is the reason behind PhysicsX $100M raise ?

The global industrial software sector witnessed a seismic validation as PhysicsX, the London-based physical AI startup, secured up to $100 million from NVIDIA—$20 million immediate investment with rights to contribute another $80 million in the next raise—just months after closing its $135 million Series B at near-$1 billion valuation.

Led by Atomico with participation from Temasek, Siemens, Applied Materials, and strategic backing from chip giant NVIDIA, this capital infusion coincides with PhysicsX quadrupling revenue over two years while serving aerospace, defense, automotive, semiconductor, and energy customers including Rio Tinto, Siemens, and Applied Materials.

This raises a crucial question: why is venture capital flowing into AI-native engineering software when traditional computer-aided engineering (CAE) platforms from ANSYS, Siemens, and Dassault commanded $12 billion annually through decades-old physics solvers, yet PhysicsX’s foundation models compress simulation timelines from weeks to hours by learning physical laws rather than numerically solving equations—fundamentally reimagining how engineers design, test, and manufacture hardware?

The $32.36 Billion Market Nobody Expected AI Could Revolutionize

The answer lies in understanding what’s happening beneath the surface of engineering software evolution.

Despite the global CAE market valued at $12.05 billion in 2025 projecting explosive growth to $32.36 billion by 2035—representing 10.1% compound annual growth—the sector faces a critical innovation paradox. Traditional simulation tools require engineers spending weeks setting up finite element analysis (FEA), computational fluid dynamics (CFD), or thermal models, then waiting days for high-performance computers solving differential equations across millions of mesh points.

Yet industries demand real-time design iterations as product complexity explodes, sustainability mandates tighten, and geopolitical pressures force sovereign manufacturing requiring faster hardware innovation cycles.

From F1 Racing to Foundation Models

PhysicsX operates a technology platform unique in industrial AI: physics foundation models that learn physical behaviors from data—compression, heat transfer, fluid dynamics, structural mechanics—enabling instant predictions without traditional numerical solvers.

Founded by Robin Tuluie (formerly Head of R&D at Renault F1 and Mercedes F1, Vehicle Technology Director at Bentley Motors) and Jacomo Corbo (formerly Chief Scientist at QuantumBlack AI by McKinsey, Chief Race Strategist at Renault F1), the 150+ person team brings motorsport’s obsession with incremental performance gains and data-driven optimization to manufacturing’s hardest problems.

Why Traditional CAE’s Solver Bottleneck Required AI-Native Alternatives

PhysicsX’s rapid funding trajectory—$170 million total with NVIDIA’s $100 million commitment—provides context for why physics AI outweighs incremental solver improvements.

When aerospace companies need designing electric jet engines or quantum chips in weeks not years, traditional simulation approaches become existential bottlenecks. Legacy CAE tools excel at accuracy but demand extensive setup time, specialized expertise, and massive compute resources running overnight batch jobs—incompatible with modern product development velocity.

NVIDIA’s investment validates PhysicsX’s approach aligns with the chip maker’s expanding interest in European startups developing technologies complementing its hardware ecosystems. Beyond immediate capital, NVIDIA retaining rights investing another $80 million signals long-term strategic collaboration—critical as GPU infrastructure becomes essential for training physics models rivaling traditional solvers.

Unlike traditional CAE requiring months learning complex software interfaces and weeks configuring each simulation, PhysicsX’s developer-first AI platform enables engineers describing what they need—component behavior under stress, thermal performance across operating conditions, fluid flow optimization—then receiving results in hours through models that learned physics from vast datasets rather than numerically solving from first principles each iteration.

The Foundation Model Architecture Behind Industrial Adoption

The funding round accelerates global expansion and fast-tracks development of larger, more powerful physics foundation models as PhysicsX scales from dozens of enterprise customers to industry-standard infrastructure.

Industry data confirms simulation software market exhibits unprecedented trajectories—$16.77 billion in 2024 reaching $60.48 billion by 2033 at 15.3% CAGR, with automotive adoption at 87% among OEMs and aerospace usage surging 47% as sectors embrace virtual prototyping.

Electric vehicle manufacturers face thermal management simulations across battery packs with thousands of cells—computationally prohibitive using conventional FEA approaches. Semiconductor fabs need optimizing chemical vapor deposition processes requiring multiphysics coupling that current CAE platforms handle inadequately.

PhysicsX differentiates through full-stack AI platform eliminating simulation complexity. The company’s technology enables rapid development of manufacturing components by simulating material behaviors and engineering designs under varied conditions—transforming pace and cost structure of manufacturing in complex industries.

Why This Matters For Global Industrial Strategy

PhysicsX’s $100 million NVIDIA commitment positions the platform within broader 2025 industrial dynamics where AI-native engineering demonstrates strategic advantages justifying massive investments.

Sovereignty and Supply Chain Transformation: Innovation in advanced manufacturing has never been more urgent or consequential. These industries underpin economies, sit central to sustainability transitions, and form heart of sovereignty challenges. PhysicsX builds into this gap with conviction that AI-native engineering software can solve fundamental challenges inherent to hardware innovation.

Simulation Economics Transformation: Traditional CAE tools require high-performance computing infrastructure costing enterprises millions annually. PhysicsX’s foundation models compress simulations from days to hours by learning physical laws rather than solving differential equations—analogous to how large language models generate text without explicitly programming grammar rules. Studies show simulation-based product testing replaced 44% of traditional prototyping efforts, while digital twins cut design errors by 49%.

Market Consolidation Acceleration: Siemens completed $10 billion Altair Engineering acquisition in March 2025, signaling consolidation wave as incumbents buy specialists creating end-to-end platforms. Yet PhysicsX’s AI-native approach differentiates from integration-focused strategies—rather than connecting existing tools, the company rebuilds simulation stack from foundation using machine learning.

The Answer: Foundation Models Meet Hardware Innovation

So why $100 million from NVIDIA for PhysicsX approaching $1 billion valuation?

Because the platform combines elements strategic buyers value: world-class founding team from Formula One bringing extreme performance optimization culture, proven enterprise traction serving critical customers like Rio Tinto and Applied Materials demonstrating technology readiness, and strategic timing where CAE market grows 10%+ annually while AI capabilities reach production-grade maturity.

The investment validates that industrial software winners emerge through physics foundation models enabling 10x speed improvements rather than incrementally optimizing legacy solvers. With aerospace companies needing electric jet engine designs in weeks, semiconductor fabs requiring real-time process optimization, and automotive OEMs demanding instant thermal simulations across battery systems, traditional CAE approaches exhaust improvement pathways.

As 72% of industrial manufacturers adopt CAE tools yet face weeks-long simulation cycles limiting design iterations, PhysicsX’s infrastructure compressing timelines to hours positions the company capturing first-mover advantages in winner-take-most dynamics.