Legal name: Secondmind Limited (formerly PROWLER.io Limited)
Secondmind Deep Dive: A Cambridge Engineering-AI Firm Whose Bayesian Active Learning Cuts CAE Simulation and Test-Bench Runs by up to 90%, with Roughly USD 73M Raised and Mazda Both Customer and Investor
A Cambridge Engineering-AI firm founded in 2016 as PROWLER.io and renamed in 2020; its active-learning software cuts simulation runs by up to 80-90 percent, with Mazda both customer and investor; roughly USD 73M raised with no round since November 2023.
Executive Summary
Secondmind Limited is a Cambridge, UK engineering-AI company founded in 2016 as PROWLER.io and rebranded to Secondmind in late 2020. Its cloud-native software, powered by Secondmind Active Learning, applies probabilistic machine learning, rooted in the Gaussian-process tradition of the Cambridge group its chief scientist co-founded, to decide which simulation or test an engineering team should run next. The product thesis is narrow and specific: every simulation and test has a cost, only a fraction of them change the engineering answer, and a data-efficient active learner can cut the number of runs needed for design-space exploration and calibration by up to eighty to ninety percent without replacing the customer's existing toolchain.
The company does not simulate physics itself; it orchestrates the tools that do. Secondmind connects to the simulation and test-bench software engineers already use through an API, an SDK or a code-free interface, and its models quantify the uncertainty behind every recommendation. This distinguishes it from solver-building peers such as Monumo, PhysicsX or Yunjing, and it is the single most important fact for positioning the company in the physical-AI landscape: it sits on the decision layer above CAE, not inside the solver.
The flagship customer relationship is Mazda, which is simultaneously a strategic investor. Mazda has collaborated with Secondmind for more than three years, joined the November 2023 funding round as an investor, and is the named reference for a fifty-nine-percent reduction in engineering hours for calibration-map creation. Other disclosed results include an electric-motor design that cut bill-of-materials cost by twenty-seven dollars per unit while improving torque, reliability and NVH, an electric-motor calibration that cut time by eighty-five percent and bench usage to a fifth, and a diesel calibration that reached minimum fuel consumption within Euro 6 limits in under four hundred test points against twenty-five hundred conventionally.
Commercial momentum in 2025 and 2026 is channel-led rather than booked-revenue-led. In July 2025 the company opened its first US presence in Detroit under Nick Appleyard, a former Siemens simulation executive, catalyzed by the signing of an unnamed major Tier 1 licensee; in June 2026 it closed a strategic reseller agreement with SC Automotive Engineering, the Sumitomo Corporation subsidiary, the first partnership SCAE has ever signed with a non-Japanese technology company. No pricing, margin or official revenue has been disclosed; a third-party estimate places annual revenue between nineteen and fifty million US dollars.
The financing base is modest and aging. The company has raised roughly seventy-three million US dollars across five instruments since 2016: a seed of about two million, a 2017 Series A of about ten million pounds, a 2019 Series B of about twenty-one point five million euros at a valuation near ninety million euros, an undisclosed Reefknot Investments addition, and the sixteen-million-dollar November 2023 round that included conversion of a Future Fund note. That round was the last one, and more than thirty months had passed as of late 2026, which is a material runway signal for a company of this scale.
The leadership is academically anchored and commercially rebuilt. Carl Edward Rasmussen, the Cambridge machine-learning professor who co-authored the standard Gaussian Processes textbook, has been chairman since 2016 and the company's first chief scientist since February 2019. Gary Brotman, an appointed chief executive since July 2021 with a serial software-operator background, leads a team that third-party estimates at fifty-one to two hundred people across Cambridge, Yokohama and Detroit.
Company Overview
Secondmind Limited was founded in 2016 in Cambridge under the name PROWLER.io, running a general-purpose machine-learning decision platform for four years before rebranding to Secondmind with a September 2020 announcement and a sharpened focus on automotive engineering. Vishal Chatrath, co-founder and chief executive from 2016 to 2021, built the company with co-founder and CTO Dongho Kim; third-party profiles additionally list Aleksi Tukiainen and Kul Singh among the founding group. Chatrath departed in January 2021 and the board appointed Gary Brotman as chief executive in July 2021, bringing in an operator whose earlier companies included QuantrolOx and VocalIQ, the latter acquired by Apple in 2015.
The scientific anchor is unusually strong for a company of this size. Carl Edward Rasmussen has served as chairman since 2016 and became the company's first chief scientist in February 2019, overseeing a research leadership team that, as of that announcement, had produced thirty-eight peer-reviewed papers and filed eight patents. The current research roster includes a senior director heading AI with a background in Bayesian optimisation and computer experiments, and principal research scientists in probabilistic modelling and Bayesian optimisation, which keeps the company's technical claim, that its active learning is engineering-grade rather than a black box, at least staffed to be credible.
The organisation is deliberately distributed along the automotive value chain. Cambridge hosts the core AI and software team, Yokohama hosts customer-facing operations for the Japanese market since January 2021 under a country manager, and Detroit, opened in July 2025, is the first US operational presence and is led by Nick Appleyard as country manager for the USA and vice-president of sales and business development for North America and Europe. Appleyard's prior career at Siemens Digital Industries Software, CD-adapco, Ricardo, General Dynamics, Ansys and Exa is itself a signal of where the company believes its buyer sits.
The mission framing has shifted with the market. The 2020-era materials framed the company as closing the gap between people and AI for decision-making under uncertainty across industries; the current site frames it as engineering AI that knows which runs matter, serving passenger-car, commercial-vehicle, off-highway and motorsport teams taking a model-based approach to systems engineering. The pivot from horizontal decision support to automotive engineering is complete, and the current materials no longer market the general-purpose decision engine.
No headcount, cash balance, burn rate, precise revenue or listing plan has been disclosed. The only quantitative personnel signals are the third-party estimate of fifty-one to two hundred employees and the composition of the public leadership and research roster.
Common Misreadings: Four Errors That Change the Analysis
The phrase Series U is a database artifact, not company vocabulary. One database labels the November 2023 sixteen-million-dollar round Series U and the master list adopts that label, while other databases call the same instrument Series B, Series B-III, a venture round or other equity. The company's own announcement describes only a sixteen-million-dollar funding round with Mazda joining existing investors, and this report treats the instrument as an extension of the Series B line rather than as a new letter. A reader who quotes Series U as a company term is repeating a data provider's convention.
Secondmind does not run the physics. The company connects to the customer's existing simulation and test-bench tools; it does not ship a solver, a mesh generator or a multiphysics engine. The eighty-to-ninety-percent run reductions are achieved by choosing better experiments, not by faster computation. Treating Secondmind as a simulation vendor overstates what it replaces and understates its integration burden.
The 2023 round was not sixteen million dollars of fresh equity alone. The announcement states the funding included conversion of a Future Fund note, so part of the instrument was previously raised convertible debt. Quoting sixteen million dollars as new cash overstates the incremental capital.
The flagship case is not an arm's-length reference. Mazda is simultaneously the flagship customer, a strategic investor and a partner of more than three years, and the fifty-nine-percent engineering-hour reduction is a company-reported figure with no published third-party audit. The result is credible and quotable, but it is not independent evidence.
Technology
The technical core is Bayesian active learning built on the Gaussian-process tradition: probabilistic surrogate models that both predict engineering outcomes and quantify their own uncertainty, used to select the next simulation or test that will most reduce that uncertainty. The company describes this as engineering-grade intelligence built for reasoning under constraints, automatically identifying the next run with explainable results and quantified confidence, and it explicitly targets teams with no machine-learning expertise.
The two productised applications are design-space exploration and calibration. Secondmind for Design Space Exploration maps high-dimensional design spaces to find feasible designs that balance competing requirements, and the company states it reduces simulations by up to eighty percent; a leading engine manufacturer reportedly discovered three times more feasible designs in half the time than with its existing simulator. Secondmind for Calibration automates experimental design for engines, e-motors and other power systems, cutting calibration time by more than fifty percent by restricting physical testing to the data that contributes most to the performance objective.
The deployment model is deliberately non-invasive. The software is cloud-native and ISO 27001 certified, provides a dedicated secure tenant per customer, offers an option to normalise proprietary data before it leaves customer servers, and integrates through an API or SDK for automation or a code-free path. That design choice is the commercial strategy: adoption does not require displacing Ansys, Siemens or Altair, which is also why the defensibility question is real.
The physics connection is indirect but genuine. Secondmind's models learn the input-output behaviour of physical simulations and physical tests, which makes them surrogate models in the engineering-AI sense, and their uncertainty estimates are what allow a run to be skipped safely. The company does not contribute solvers, meshers or physics formulations; its contribution is in experiment selection under uncertainty, which places it as a peer of active-learning and surrogate vendors rather than of simulation vendors.
The research base is the moat, and it is narrow. Thirty-eight peer-reviewed papers and eight filed patents as of the February 2019 chief-scientist announcement, a chairman who wrote the standard textbook on Gaussian processes, and a research team led from Bayesian optimisation give the company genuine depth in its chosen method. The limitation is equally clear: the method class is mature, well published and replicated by competitors, so the defensibility comes from engineering packaging and customer trust rather than from an algorithmic secret.
Product and Commercial Model
The commercial model is a software subscription sold to OEMs, full-service suppliers and Tier 1 suppliers in passenger car, commercial vehicle, off-highway and motorsport, with the value proposition measured in engineering hours, test-bench occupancy and simulation runs saved.
The disclosed traction is named relationships and channel agreements, not booked revenue. Mazda is the flagship, an investor and a partner of more than three years; an unnamed major Tier 1 supplier signed as a licensee in 2025 and catalysed the Detroit opening; SCAE, the Sumitomo Corporation subsidiary, became a value-added reseller in June 2026, its first agreement with a non-Japanese technology company. No pricing, gross margin, renewal or official revenue figure has been published.
The revenue estimate is third-party and wide. One database estimates annual revenue between nineteen and fifty million US dollars, a range so wide that it should be read as an order-of-magnitude signal rather than a measurement, and it sits alongside a pre-Series-A-stage revenue posture that the same source reports inconsistently.
The channel shift is the most consequential commercial fact of 2026. A reseller agreement puts local engineering credibility, in Japan's case SCAE's, between Secondmind and the buyer, which lowers adoption friction in a market where incumbents' local presence is deep. It also caps margin, and it makes the company's growth partly dependent on a small number of channel partners.
The buyer-overlap risk is structural. Siemens, Ansys, Altair and Dassault all sell to the same engineering organisations and are all adding AI-assisted exploration and calibration to their own suites; a software partner that never displaces the incumbent suite is also a software partner the incumbent suite can absorb.
Competitive Landscape
Secondmind occupies the decision layer above CAE, and its competitors are defined by which layer they attack.
Against solver-building AI natives such as Monumo, PhysicsX and Yunjing, the comparison is architecture against decision. Those companies generate or accelerate the physics; Secondmind reduces how often physics must be generated. The two layers are complementary more than substitutive, which is why partnership, not head-on competition, is the likelier pattern.
Against simulation incumbents adding AI, such as SimScale with its NAVASTO integration and the majors' own AI features, the comparison is the hardest one. NAVASTO, the closest structural analogue to Secondmind in design-space exploration, was acquired by Autodesk, which is the clearest available signal of where standalone active-learning independents end up. Incumbents have the distribution, the existing data pipelines and the engineering relationships; Secondmind has focus and neutrality.
Against Neural Concept, the comparison is one of method against scope. Neural Concept sells geometry-driven surrogate prediction into the same OEM buyers with a far larger war chest; Secondmind's defence is that calibration and test reduction require physical experimentation, where its active-learning framing is more directly applicable than shape-based prediction.
Against in-house OEM teams, the comparison is a de facto one. Mazda's own advanced-AI research and education strategy, which the partnership explicitly supports, is also the capability that could eventually internalise what Secondmind sells, and every strategic investor carries that dual-use risk.
Against other active-learning and experimental-design vendors, the field is thin but real, and the 2025 Autodesk acquisition of NAVASTO has already consolidated the most comparable independent. Secondmind's remaining differentiation is automotive-specific packaging, uncertainty quantification as a product feature, and a channel position in Japan.
Financing
The seed in September 2016 was about two million dollars from angels including Amadeus Capital Partners, Flourish Ventures, IA Capital Group, Passion Capital and Andy Bechtolsheim, under the PROWLER.io name.
The Series A in July 2017 was about ten million pounds, led by Cambridge Innovation Capital and Atlantic Bridge with existing investors including Passion Capital, Amadeus and SGInnovate.
The Series B in May 2019 was about twenty-one point five million euros, at a valuation of roughly ninety million euros, from Amadeus, Atlantic Bridge, Cambridge Innovation Capital, Mandatum Life, Passion, Pearson, RB Capital, SGInnovate and Tencent, with Reefknot Investments, the Temasek and Kuehne plus Nagel venture vehicle, adding to the round in October 2019 in an undisclosed amount.
The November 2023 round was sixteen million dollars and included conversion of a Future Fund note, with Mazda joining as a strategic investor alongside Amadeus, Atlantic Bridge and Cambridge Innovation Capital. The round was announced on 27 November 2023 together with the expanded Mazda partnership, and it was explicitly earmarked for sales and marketing scale-up and acceleration of the cloud-native optimisation engine.
Cumulative funding is roughly seventy-three million dollars across five instruments by one database's count and seventy-nine point one million by another's, and the difference is explained by how the Reefknot addition and the note conversion are aggregated. The more actionable fact is the gap: no round has been disclosed since November 2023, a window of more than thirty months as of late 2026, which for a company whose own 2023 announcement framed the raise as a scale-up step is a runway signal that a reader quoting only the total will miss.
DISCREPANCIES
1. November 2023 round label. Convention A: Series U (one database, adopted by the master list). Convention B: Series B (one database), Series B-III (another), venture round or other equity (a third). Convention C: no series letter at all, described only as a sixteen-million-dollar funding round (company announcement). Ruling: the company announcement is authoritative for the instrument's existence and amount; the letter labels are data-provider conventions. This report treats it as an extension of the Series B line and does not use Series U as a company term.
2. Total capital raised. Convention A: about seventy-three million dollars across five rounds (one database, adopted by the master list). Convention B: seventy-nine point one million (another database). Ruling: both are database aggregates with different treatment of the Reefknot addition and the note conversion; the honest statement is roughly seventy-three to seventy-nine million dollars total, and neither figure is a company disclosure.
3. Founders. Convention A: Vishal Chatrath and Dongho Kim as co-founders (company statements and business press). Convention B: the same two plus Aleksi Tukiainen and Kul Singh (one database). Convention C: founded by chief executive Gary Brotman (one startup database). Ruling: A is company-confirmed and adopted; B is reported as a database-level expansion of the founding group; C is incorrect and is recorded as an error, because Brotman was appointed by the board in July 2021, six years after founding.
4. Chief executive. Convention A: Vishal Chatrath, co-founder and chief executive, 2016 to 2021 (company and press). Convention B: Gary Brotman, chief executive from July 2021 (company announcement and all current materials). Ruling: both are correct for their periods; the transition is documented, with Chatrath's departure reported in January 2021 and Brotman's appointment in July 2021, and no source documents an interim period.
5. Headcount. Convention A: fifty-one to two hundred employees (two databases). Convention B: more than ten thousand (one database). Ruling: A is adopted as a plausible range for a company of this funding profile; B is treated as a data error and is not used anywhere in this report.
6. Series B valuation. Convention A: about ninety million euros (one database). Convention B: one hundred million dollars post-money (another database). Ruling: the two are consistent after currency conversion at 2019 rates; the report states roughly one hundred million dollars at the May 2019 Series B.
7. Rebrand timing. Convention A: announced 29 September 2020 (company release via Business Wire). Convention B: rebranded in October 2020 (investor databases). Ruling: the announcement date is authoritative; October is recorded as the effective rebrand month by databases, and the report uses September 2020 for the announcement.
8. Revenue. Convention A: annual revenue estimated at nineteen to fifty million dollars (one database). Convention B: no official revenue disclosure (absence of company disclosure). Ruling: the estimate is reported as a third-party figure with its width intact, and the company is treated as not having disclosed revenue.
Risks and Due Diligence Gaps
The longest gap is independent validation of the headline percentages. The eighty-percent simulation reduction, the ninety-percent figure for design-space exploration, the fifty-nine-percent Mazda engineering-hour reduction and the twenty-seven-dollar bill-of-materials reduction are all company-reported, and the Mazda case is additionally a strategic-investor relationship. Due-diligence action: reference calls with at least two non-investor licensees and access to measured before-and-after engineering-hour data.
The financing runway is the second gap. No disclosed round since November 2023, a scale-up framed raise, and a company whose commercial build-out in 2025 and 2026, Detroit, a channel in Japan, senior sales hires, implies continued burn. A reader who quotes the seventy-three-million-dollar total without the thirty-month gap will misread the balance sheet.
The incumbent-absorption risk is structural, and the market has already shown the pattern. NAVASTO, the closest independent analogue in design-space exploration, was acquired by Autodesk. Secondmind's neutrality is its pitch and also its exposure: the same buyer set is served by suites that can bundle a comparable feature, and a partner that never displaces the suite can be displaced by the suite.
The channel concentration is new and untested at scale. SCAE is the first reseller and the first partnership of its kind for that firm; the US push rests on one unnamed Tier 1 licensee. Two channels carrying the growth plan, with no disclosed renewal or expansion data, is a thin base for a company claiming a widening pipeline.
The method class is mature and replicable. Gaussian-process active learning is well published, and the company's own research leadership is the differentiator rather than any exclusive algorithm. Competitors with larger engineering teams can package equivalent functionality, and the moat is packaging, trust and uncertainty presentation rather than a proprietary core.
Team scale is small against the buyer set. Fifty-one to two hundred people against Siemens, Ansys, Altair and Dassault organisations of thousands means delivery and support capacity is the binding constraint on how many concurrent OEM programmes the company can carry, and the 2025 to 2026 expansion into a third country increases that strain.
Overall Assessment
- Technology 7.0. A genuinely deep Bayesian active-learning stack anchored by a Cambridge chair who co-authored the standard Gaussian-processes textbook, productised into design-space exploration and calibration with quantified uncertainty and explainable next-run selection. Deducted points: the method class is mature and published, the performance claims are company-reported with no third-party reproduction, and the company contributes no solver or physics asset of its own.
- Commercialisation 6.5. A flagship customer of Mazda's standing who is also an investor, an unnamed major Tier 1 licensee catalysing a Detroit opening, a first-of-its-kind Japanese reseller in SCAE, and a third-party revenue estimate of nineteen to fifty million dollars. Deducted points: no official revenue, pricing, margin or renewal disclosure, traction that is relationships and channels rather than booked contracts, and a revenue range too wide to plan against.
- Financing 6.0. Roughly seventy-three to seventy-nine million dollars raised across five instruments from a credible investor base including Tencent, Amadeus, Atlantic Bridge, Cambridge Innovation Capital, Reefknot and Mazda. Deducted points: the last round was November 2023 and included a note conversion, more than thirty months have passed with no disclosed successor, and the 2019 valuation of roughly one hundred million dollars leaves limited headroom if the next round is flat.
- Product maturity 7.0. A delivered, ISO 27001 certified, cloud-native SaaS with two productised applications, API, SDK and code-free paths, per-customer tenants and a data-normalisation option, deployed into real engine, e-motor and diesel programmes. Deducted points: no published customer deployment success rate, no third-party benchmark of the headline percentages, and a product whose value is realised only when integrated with the customer's existing toolchain.
- Market position 6.5. A focused independent in the decision layer above CAE, with a rare combination of academic depth, automotive-specific packaging, a strategic OEM investor and a Japanese channel. Deducted points: direct exposure to incumbents who serve the same buyers and can bundle the feature, a market that has already absorbed the closest analogue through acquisition, a small team against that buyer set, and two unproven channels carrying the growth plan.
References
- Secondmind company site, including the Secondmind Active Learning framing, the design-space-exploration and calibration product claims, the case results for e-motor design, ICE calibration, e-motor calibration, piston-bowl design, diesel calibration and e-powertrain calibration, and the Mazda quotation from Tomohiko Adachi
- Secondmind announcement of the expanded Mazda strategic partnership and the sixteen-million-dollar round, 27 November 2023, including Mazda joining as investor with Amadeus, Atlantic Bridge and Cambridge Innovation Capital, the Future Fund note conversion, and the quotes from Gary Brotman, Naohito Saga, Hermann Hauser, Kevin Dillon and Ian Lane
- Secondmind announcement of Nick Appleyard as country manager USA and vice-president of sales and business development, 8 July 2025, including the Detroit presence, the major Tier 1 licensee signing and Appleyard's Siemens, CD-adapco, Ricardo, General Dynamics, Ansys and Exa background
- Secondmind and SCAE strategic reseller agreement announcement, 9 June 2026, including SCAE's first partnership with a non-Japanese technology company, the Sumitomo Corporation subsidiary context and the Toshimi Yamanoi quotation
- Secondmind chief-scientist announcement, February 2019, including Carl Edward Rasmussen as chairman since 2016 and first chief scientist, the thirty-eight papers and eight patents, and the Vishal Chatrath quotation
- Business Wire rebrand announcement, 29 September 2020, including four years as PROWLER.io and the Secondmind Decision Engine launch
- Funding database profiles including the seed, Series A, Series B and November 2023 rounds, the seventy-three to seventy-nine million dollar totals, the ninety-million-euro and one-hundred-million-dollar 2019 valuations, the employee estimate of fifty-one to two hundred and the revenue estimate of nineteen to fifty million dollars
- Business Weekly coverage of Vishal Chatrath's departure, January 2021, and the board appointment of Gary Brotman as chief executive, July 2021
- Business Weekly Awards 2026 shortlist coverage, including the calibration-time and simulation-run reduction claims and the Detroit and SCAE expansion notes
This report is compiled from public sources. It is for reference only and does not constitute investment advice. Data is current as of October 24, 2026; please refer to official company announcements and filings.
Related papers9
- Transolver: A Transformer Solver for Large-Scale Industrial Geometries
- Fourier Neural Operator (Original Paper)
- DeepONet (Original Paper)
- KATOsuper: Surrogate-accelerated neural topology optimization with sensitivity-consistent Fourier neural operators
- Deep-Learning Solvers and Surrogates for Infinity and p-Laplace Problems
- DeepFEAv2: Deep Learning for Transient Finite Element Analysis Beyond Structured Meshes
- Sparse-Observation Atmospheric Thermal Forecasting with Physics-Informed Neural Networks for Climate-Aware Digital Twins
- Analysis of trade-offs in urban heat mitigation using a Bayesian Optimization framework for an urban canopy layer model
- A Fully Implicit Electromagnetic Energy- and Charge-Conserving PIC Model for High-β Plasma Simulations
Related cases
- DrivAerML High-Fidelity Road-Car Aerodynamics Dataset and Surrogates
- AirfRANS High-Fidelity Airfoil Surrogate Benchmark Case
- Physics-Informed CNN for Battery Pack Temperature Fields
- IASC-ASCE Structural Health Monitoring Benchmark and Physics-Informed Damage Identification
- Wind Turbine Aeroelastic Simulation and Wake Control: A Verifiable Chain from OpenFAST to Reinforcement Learning
Related reports66
- 51World Deep Dive: A Listed Digital Twin Company Where 51Sim Becomes the First Growth Engine While 51Earth Nearly Vanishes, and Two Irreconcilable Financing Naming Conventions Must Both Be Kept
- AMI Labs: A USD 1.03B Seed Bet on World Models, LeCun's Non-Generative Path and Its Blank Scorecard
- Anshi Pacific (PERA) Deep Dive: A Beijing Domestic-CAE Leader Founded in 2003 Whose PERA SIM General Simulation, AIROM Reduced-Order Models, AgentX Multi-Agent and Jingzhi iGPT/Whale Span from Industrial Simulation to Physics AI
- AspenTech Deep Dive: The Process-Simulation Incumbent Acquired by Emerson and Its Hybrid Industrial-AI Route
- Autodesk Deep Dive: Neural CAD Foundation Models and a $200M World Labs Bet on Physical-World AI, with NAVASTO Adding CFD Surrogate Capability
- Bentley Systems Deep Dive: iTwin Digital Twin and Generative AI Empower Full-Lifecycle Infrastructure Engineering
- BeyondMath Deep Dive: A Foundation Physics AI on the Neural-Operator Route, a $19M Consortium Budget, and Three Misread Facts
- Cadence Deep Dive: System-Level Multiphysics EDA and the Millennium M2000 Supercomputer Anchor Physical AI, with AuraStack Agentic AI and Fidelity GPU CFD as Key Levers
- COMSOL Deep Dive: Four Decades of Multiphysics Solvers, Entering Physical AI via DNN Surrogate-Model Training and All-Physics GPU Acceleration
- Dassault Systèmes Deep Dive: 3DEXPERIENCE and SIMULIA Virtual Twin Physics Behavior Anchor Industrial AI, with LEO Virtual Companion and GPU Explicit Dynamics as Physical-AI Levers
- DP Technology: From the Gordon Bell Prize to National-Level Unicorn Status and the Door of a Hong Kong Listing, the AI-for-Science Leader's Technical Delivery and Commercial Blank
- Emmi AI: From Linz Spin-out to a EUR 330M Mistral Acquisition in 17 Months
- Fainite Deep Dive: A Zurich Physics-AI Simulation Startup Whose Caltech, ETH Zurich, Cambridge and Google Team Uses Physics-Aware Deep Learning to Cut Simulation from Tens of Hours to Seconds, With Only Venture Kick Grants and No Priced Round
- Fysics AI: The PhysX Founder's Second Act, a Domestic Differentiable Physics Engine and Its Ecosystem Play
- Frontop (301313) Deep Dive: From Digital Exhibition Halls to a Physical-AI Data Foundation - FTE Engine, FunSim Simulation and the Gap Between Positioning and Revenue
- Godela Deep Dive: A YC X25 Physics-AI Simulation Startup Whose Ex-Apple, Google and Intel Team Uses Physics-Constrained Surrogate Models to Cut Simulation from Days to Seconds, at Seed Stage with Pilot-Only Access
- HollyWave (688682) Deep Dive: RDSim AI Electromagnetic Simulation and Engineering Agents Close the Simulation-Measurement-Product Loop
- IAT (Alt) Deep Dive: China's First Listed Independent Automotive Design Company Pivots to AI-native R&D with the Yuanfang, Yufeng and Taiyi 3.0 Product System and an NVIDIA Omniverse Partnership, Reporting RMB 975M Revenue and a Widened RMB 250M Net Loss in 2025
- Inductiva Deep Dive: A Closed Case in Open-Solver Orchestration, Eight Premises That Must Be Corrected, and Where Co-Founder and CTO Hugo Penedones Went Next
- Intesim (INTESIM) Deep Dive: A Home-grown Chinese CAE Leader Founded in 2009 with Self-developed Solvers for Structure, Fluid, Electromagnetics, Thermal and Acoustics and Multiphysics Coupling, a ~RMB 100M B Round from CNII and Shenzhen Capital, and 2025 Revenue and Bookings Both above RMB 100M
- iSoftStone (301236) Deep Dive: A ChiNext-Listed IT Services Giant Whose TianShu Platform Builds a Three-Layer Physical-AI Stack with Self-Developed TwinHub Digital Twin, FSim Industrial Simulation, CAE Multiphysics and Phyxis Embodied Control
- Jua Deep Dive: The Earth Physics Transformer as a Large Physics Model for Energy Trading
- Keysight (with ESI Group) Deep Dive: Virtual Prototyping and Hybrid AI Bridge Measurement and Simulation, with Natural-Language Assistants in EDA
- Lightwheel Deep Dive: A Physical AI Data and Evaluation Infrastructure Startup Founded by a Former NVIDIA Simulation Lead, Seven Rounds Totalling Roughly CNY 2.65 Billion at a Valuation Above USD 2 Billion, Seats on Both the Newton and EgoVerse International Standards Committees, With Two Incompatible Accounts of Q1 Orders
- Luminary Cloud: The Physics AI Model Factory Backed by NVentures
- MathWorks Deep Dive: MATLAB/Simulink and Simscape Multiphysics Power Physical AI, with R2026a Copilots and an Agentic Toolkit for Edge Intelligence
- Maniformer (Mifeng) Deep Dive: An Agibot-Spun Physical-AI Data Infrastructure Firm Whose Bodyless MEgo Capture Hardware Is in Mass Production and Which Closed Three Rounds in Half a Year Led by China Telecom
- Mistral AI: How Physics AI Moved From Language Models to EUV Lithography and Nuclear Power
- Monolith AI: The British Engineering-AI Firm That Cut Physical Test Counts Instead of Accelerating Simulation, Now Inside CoreWeave
- Monumo Deep Dive: A Cambridge Deeptech Company Redesigning Electric Motors with the Anser AI Engine, Tens of Millions of Multiphysics Simulations per Day, GBP 10.5M Raised in Total, and a Series A Being Raised
- NAVASTO: Moving Solver Insight Onto the Designer's Desk, and How Autodesk Bought That Entry Point
- Navier: Handing CFD Workflows to Agents, a Seed-Stage Bet on AI for Hardware Engineering
- Neural Concept: An EPFL Spin-out Betting on Neural Surrogates, with $138M Raised
- NP Co. (NP Company) Deep Dive: A 2025-founded Paris Inria-born AI-native simulation startup that raised a EUR 6M pre-seed led by Partech with Mistral AI co-founders as angels, claiming a pre-trained physics foundation model at 1,000x speedup — most claims unverified
- NVIDIA Deep Dive: Building the Full-Stack Physical-AI Infrastructure with PhysicsNeMo, Warp, Cosmos World Foundation Models and Omniverse
- Physical Intelligence Deep Dive: A $2.1B Bet on the GPT Moment for Robotics, With pi0.7 Self-Reporting 60-80 Percent Success on Unseen Tasks and Neither Pricing nor Revenue Disclosed for the Pi API
- Physical Superintelligence Deep Dive: A $58M Seed Bet on Virtual Physicists, With Emmy Targeting AI Data Center Multiphysics Optimization as Its First Commercial Proof Point
- PhysicsX: Pretrained Geometry and Physics for Industrial Engineering at a $2.4B Mark
- Prometheus: A $41B Mark, $18.2B Raised, and Not Yet a Product
- PTC Deep Dive: Creo/Onshape and the ThingWorx Digital Thread Power Physical AI, with NVIDIA Isaac Sim Bridging Robotics Design to Simulation
- Qianjue Technology Deep Dive: How a Body-Agnostic Embodied Brain Pairing the Predictive World Model Polibrain with Brain-Inspired Partitioning and Polynomial Representation Reached a Hundred Thousand Cleaning and Service Robots
- Quanscient Deep Dive: Cloud-Native and Quantum-Powered Multiphysics Simulation as the Data Engine for AI-Native Hardware Engineering
- Rescale Deep Dive: An HPC Operating System and Desktop Engineering Workbench in the Cloud, a $115M Series D, and Five Misread Facts
- Schrödinger Deep Dive: Physics-Based FEP+ Simulation and Machine Learning Power Molecular Discovery, with the 2026 Bunsen Agentic Co-Scientist
- Shichuang Technology Deep Dive: SupreCAST Cloud-Native Material-Processing Simulation and Industrial Agents Advance the CAE+AI Strategy
- ShonCloud (Shunyun) Deep Dive: Meshless Particle-Method CFD for EV and Nuclear Fluid Multiphysics
- Shuzhi Maijin Deep Dive: A Shenzhen Fully Self-Developed SNN-plus-LBM CFD Startup Claiming Second-Level Results on Hundreds-of-Millions of Grids on Consumer Hardware, With a Tens-of-Millions-RMB Angel Round From a Ministry-of-Finance-Affiliated Fund and an FPGA Migration Plan
- Siemens (incl. Altair) Deep Dive: The 2025 Altair Acquisition Unites Simcenter and Altair, with PhysicsAI Geometric Deep Learning and PhysicsAI Generate as the Physical-AI Core
- Silurian Deep Dive: A Generative Forecasting Transformer Earth-System Foundation Model for Weather Risk on Energy and Infrastructure Assets
- SimScale Deep Dive: A Cloud CAE Platform Built on Core-Hour Subscriptions, EUR 52M of Equity Kept Separate from EUR 25M of Debt, and Five Misread Facts
- Skild AI Deep Dive: A General Robot Foundation Model Incubated at CMU's Robotics Institute, With a USD 1.4B Series C at a USD 14B Valuation, a Conflicting Series B Size, and an ARR Figure That Is a Run-Rate Rather Than Revenue
- Songying Deep Dive: A Huawei-Cloud-Alumni Startup Building a Domestic Physical AI Operating System, With ORCA OS Positioning Against NVIDIA Omniverse, Three Incompatible Financing Narratives, and a Verifiable Third-Party Anchor in a CNY 20 Million Minority Investment
- Sudo Tech Deep Dive: Zero-Real-Data Simulation Training, the Fully In-House Sudo R1 Stack, and an Atomic-Skill Commercialisation Strategy from a Startup Valued Above CNY 20 Billion in 17 Months
- Suochen Technology: The Pain of Becoming a Physics-AI Platform from a CAE Vendor
- Synopsys (Ansys) Deep Dive: The 2025 Ansys Acquisition Turns the EDA Leader into a Silicon-to-Systems Engineering Powerhouse, with Ansys SimAI and the NVIDIA Physical-AI Partnership as the Core Physics-AI Levers
- TenFong Deep Dive: An Academician-Led Chinese Industrial Simulation Software Vendor Whose Official Timeline and Database Records Use Incompatible Round Names, With an Official Financing PDF and an Investment Database Both on File
- Tianfu (Nanjing Tianfu Software) Deep Dive: A 2011-founded Chinese Intelligent-CAE Veteran That Closed a Several-Hundred-Million-RMB Series C, Pivoted to an Industrial World Model in 2024, With Unreconcilable Financing and Valuation Records and No Disclosed Revenue
- Tongyuan Soft Control Deep Dive: MWORKS (Syslab/Sysplorer) as a Domestic MATLAB/Simulink Alternative, with MWORKS AI and the MOGICK Engineering Agent Advancing Home-Grown Physical AI
- Vinci AI: Turning Physics Simulation from a Checkpoint into a Continuous Constraint
- Westlake Robotics Deep Dive: Westlake University's First Robotics Spin-out Builds a Dual-Pretrain Brain-and-Cerebellum Architecture with WR1 Whole-Body Foundation Model and GAE Behaviour Expert
- World Labs: Fei-Fei Li's Spatial-Intelligence Unicorn, Acquired by AMD for $8.2B
- XVERSE Deep Dive: A Simulation-World Route to Physical AI Built on the XVERSE Model Family, an Edge-Cloud 3D Engine and the Physics-Based MotionGen Model
- SimWE: From In-House Solvers to a Physical-AI Engine, the Domestic CAE Leader and What Comes Next
- Yunjing Zhifang Deep Dive: A Peking-University-Incubated Cloud-Native GPU CAE Company with the Simversus Platform and the YJ-DORGP Small-Sample Intelligent Solver, Reporting 275 Percent Order Growth, 160 Percent Revenue Growth and Breakeven in 2025
- Zhongke Diwiji (Five Ages) Deep Dive: A CAS-Institute-of-Automation and Tsinghua Spin-out Whose FAM Embodied Foundation Model Uses Action Optical Flow for Cross-Embodiment Generalisation, Reaching 97% Success from 3 to 5 Demos, with EraBrain and Jiling Robots Serving Sinopec, Leapmotor and Midea
- ZWSOFT (688083) Deep Dive: Overdrive Kernel and ZW3D Wukong 2027 Underpin an Integrated CAD/CAE/CAM Stack, with a CAx Agent Platform as the Practical Route to Engineering AI
This report is compiled from public sources. It is for reference only and does not constitute investment advice. Data is current as of the publication date; please refer to the company’s official filings.