Legal name: Suzhou ShonCloud Technology Co., Ltd.
ShonCloud (Shunyun) Deep Dive: Meshless Particle-Method CFD for EV and Nuclear Fluid Multiphysics
ShonCloud builds a fluid-multiphysics stack (shonDy particle method, shonTA thermal, shonFlow CFD) for EV and nuclear, registered in Suzhou; closed a ~RMB 100M Series B in March 2025, with AI still an early overlay.
Executive Summary
ShonCloud Technology, formally Suzhou ShonCloud Engineering Software Co., Ltd., was founded on June 26, 2019 and is registered in Xiangcheng district, Suzhou, with Wang Xi as legal representative. It builds fluid and thermal management multiphysics simulation software. The product line published on the company website comprises four items: shonDy for particle-based fluid simulation, shonTA for multiscale thermal analysis, shonFlow for general computational fluid dynamics, and shonMesh for meshing and pre-processing. The 2026 campus recruitment material additionally names shonEM for electric motor design and shonGear for drivetrain design, two products the website does not currently list. The registered entity and main research base sit in Suzhou, with subsidiaries in Shenzhen and Beijing, and third-party company profiles also mention a research office in Karlsruhe, Germany.
Public sources record four funding rounds. A Pre-A round on December 23, 2020 raised the tens of millions of renminbi from Shunrong Capital and the Suzhou Xiangcheng fund, described as an angel round in the account of the local state-owned investment platform; an A round on January 28, 2022 raised the tens of millions of renminbi from Suzhou Guofa Venture Capital, Suzhou Asset Management and Xiangcheng state capital; an A+ round on December 28, 2023 came from Oriza Holdings with the amount undisclosed; and a B round on March 19, 2025 raised nearly 100 million renminbi from the Industrial Machine Tool Industry Investment Fund, Oriza Holdings, Shenzhen Hongtu Chongshi and Shenzhen Capital Group. A separate incubator brief describes the B round as a joint investment by the Industrial Machine Tool fund and Hongtu Chongshi, a subset of that four-party list. A cumulative figure of 1.40 hundred million renminbi appears only in an AI-generated summary on an investment portal and is recorded here as an unverified estimate.
More unusually, its 2026 campus recruitment material discloses a revenue level: six years of operation, a B round of nearly 100 million renminbi, and annual revenue in the tens of millions of renminbi. The same sentence appears in a July 7, 2025 district media release, where the completed round is named as A+; since the B round closed in March 2025, the recruitment wording lags the most recent round and cannot serve as evidence of the current financing stage. The company gives no precise figure or reference year, and registry financials for 2023 show revenue of 12.67 million renminbi, an order of magnitude below the stated tens of millions.
Headcount appears in three versions that this report keeps side by side: business registry records for 2025 list 34 insured employees, a recruitment platform lists 50 to 99 people, and an encyclopedia entry says more than 40. A July 2025 district media release states that over 70 percent of staff hold a master's degree or above, with 11 full-time PhDs, 2 national-level talents, 8 provincial or municipal talents and 4 foreign experts; the 2026 recruitment material keeps the 70 percent masters share and the 4 foreign experts but lists 1 national-level talent and 6 provincial or municipal talents. Customers include Huawei, FAW, SAIC, BYD, CNNC, AVIC, American Axle and Qingshan Industries, with university partners including the Naval University of Engineering, Chongqing University of Technology, Beijing Institute of Technology and Shenzhen University.
In physical AI terms the position is physical AI for free surfaces and moving boundaries. The technical core is the particle method, specifically least squares moving particle semi-implicit, and its defining advantage is that no mesh is required. That removes a fundamental constraint, since mesh generation often consumes more than half of a simulation workflow on complex geometry, and moving boundaries such as rotating gears, splashing oil and breaking droplets distort meshes continuously and force repeated remeshing. The particle method discretizes fluid into particles carrying physical quantities, so boundary motion is just particle movement, which suits these problems naturally.
The commercial landing point is specific: oil churning loss and oil shedding simulation in electric drive systems. At high rotational speeds, lubricating oil is agitated by gears, thrown onto housing walls and forms films and mist, determining lubrication efficiency, heat removal and power loss. Mesh-based methods handle this combination of free surface, high-speed rotation and large deformation at very high cost, while the particle method is the more natural choice.
One framing point matters. This is a company with a precise technology choice, still-small scale, and a position near the breakeven line. Set registry revenue of 12.67 million renminbi for 2023 and 34 insured employees in the same year beside the self-reported revenue in the tens of millions of renminbi, revenue per employee cannot be pinned to a single figure; the published data spans from a few hundred thousand renminbi to over one million. Its official product line covers four products, and the maturity and market share of each cannot be judged from public information. The key question is not the technology narrative but whether the oil churning niche can support a company of meaningful scale.
Company Overview
Origins and Technical Route
ShonCloud was founded in 2019 by Wang Xi, whose academic background in particle numerical methods, high performance computing and multiphysics coupling maps directly onto the company's technical direction. The stated mission is advanced simulation serving engineering innovation, with positioning as a next-generation fluid multiphysics simulation platform.
The word next-generation usually means two things in Chinese CAE: an algorithm route different from the conventional one, here particle rather than mesh methods, and an architecture different from the conventional one, here GPU-native and cloud-ready. ShonCloud has both, which differentiates it from traditional domestic CFD vendors.
Its development path also differs from most domestic CAE companies. Most start by replicating the general functionality of foreign tools, building a general structural or fluid solver and then adding modules. ShonCloud started from one specific method, particle simulation, established a foothold in scenarios where conventional tools perform poorly such as splashing, oil churning and fluid-structure interaction, and then expanded into general CFD through shonFlow and system-level thermal management through shonTA. The advantage is clear differentiation and limited competition, and the disadvantage is a small initial market that forces confrontation with mature rivals on expansion.
Organization and Talent
The registered entity and main research base sit in Xiangcheng district, Suzhou, with subsidiaries in Shenzhen and Beijing. Third-party company profiles additionally mention a research office in Karlsruhe, Germany. Karlsruhe has a technical rationale, being a German center for fluid mechanics and numerical computation with accumulated work in CFD and particle methods at the Karlsruhe Institute of Technology, which helps recruit European talent and support international collaboration.
The talent structure is the most notable asset. The company-reported figures of 11 full-time PhDs, 2 national-level talents, 8 provincial or municipal talents and 4 foreign experts, set against a registry count of 34 insured employees, describe a dense senior team, which explains both the ability to build a particle solver in house, since that work demands deep numerical method foundations, and the high share of personnel cost. The 2026 recruitment material reports fewer national and provincial talents, and public sources do not reconcile the two versions.
The 2026 campus recruitment list reveals strategy: fluid numerical algorithm engineer, high performance parallel algorithm engineer, physical AI algorithm engineer, electromagnetic numerical algorithm engineer, mesh algorithm engineer and graphical interaction software engineer. The physical AI role is an explicit signal that surrogate models and data-driven methods are entering the product roadmap.
Customers and Market
The disclosed customer list spans automotive, nuclear, aerospace, marine, electronics, general machinery and appliances, including Huawei, FAW, SAIC, BYD, CNNC, AVIC, American Axle and Qingshan Industries, with university customers including the Naval University of Engineering, Chongqing University of Technology, Beijing Institute of Technology and Shenzhen University.
Customer counts come in three versions: the company says hundreds, third-party material says about a hundred domestic and international customers, and conference material says close to a hundred. The differences are noted in the discrepancies. What can be confirmed is coverage of the main players in the electric drive chain including Huawei, BYD, FAW, SAIC and Qingshan Industries, plus nuclear through CNNC, two domains where fluid and thermal simulation demand is most rigid.
Technology
Why Particle Methods Suit Splashing and Oil Churning
Understanding the technical value requires understanding why mesh-based methods struggle with free surfaces.
Mesh-based CFD, mostly finite volume, discretizes the domain into cells and describes free surfaces through volume of fluid or level set. The approach performs well on most engineering problems but becomes very expensive in three situations. Violent free surface deformation, where droplet breakup, splashing and wave breaking make interfaces cross cells repeatedly and require very fine meshes plus special interface reconstruction. Large boundary displacement, where rotating gears, reciprocating pistons and deforming vessels force mesh distortion and remeshing, each remeshing introducing interpolation error and cost. And large solid deformation with fluid-structure coupling, where structural deformation twists the mesh and coupling iterations diverge easily.
Particle methods including smoothed particle hydrodynamics, moving particle semi-implicit and their variants discretize fluid into particles carrying mass, momentum and energy, approximating field variables with kernel functions and requiring no mesh. Free surfaces are represented naturally by the presence or absence of particles, so breakup and splashing are just particle separation. Moving boundaries only require moving boundary particles, and large solid deformation can use particles or hybrid particle-finite-element approaches. The cost is harder control of accuracy and conservation, since uneven particle distribution introduces error, plus high computational load from neighbor interactions and complex boundary condition treatment.
The least squares moving particle semi-implicit method used by ShonCloud is an improved variant of moving particle semi-implicit that applies least squares spatial discretization to raise accuracy and stability and alleviate the pressure oscillation of the classical method. That choice indicates real numerical research rather than simply wrapping an open-source SPH code.
Oil Churning: An Underestimated Engineering Problem
Oil churning loss in electric drive systems is the clearest commercial landing point and deserves separate explanation.
Electric vehicles use oil cooling and lubrication in motors and reducers, with oil agitated by gears and thrown onto housings and bearings to form films and mist that both lubricate and remove heat. The resulting power loss can represent a meaningful share of total drivetrain loss at high speed and directly affects range. Reducing it requires optimizing oil level, gear geometry, housing flow guides and oil passages, all of which require knowing where the oil is, how fast it moves and how thick the film is.
Conventional approaches are dynamometer testing, which is costly, slow and yields only aggregate numbers without detail, or mesh-based CFD, which involves complex geometry, many moving boundaries and heavy meshes with runs lasting days. Particle methods have a natural advantage because oil splashing, breakup and adhesion are all free surface phenomena and gear rotation is just boundary particle motion without remeshing. GPU parallelism further brings runtime into an acceptable range.
The commercial value lies in the problem being painful enough to matter for range, narrow enough to have limited players, and hard enough to constitute a barrier. Narrow means a low ceiling and hard means a high barrier. ShonCloud's strategy is to build technical credibility in this scenario and then extend to adjacent ones such as ship wave impact, tank sloshing and pump and valve internal flow.
Division of Labor Across Three Solvers
shonDy is the particle-based fluid solver, meshless and strong in splashing and fluid-structure interaction, supporting GPU parallelism and multi-GPU and multi-CPU distributed computing with integrated oil shedding and fluid breakup features. It targets free surface problems conventional tools handle poorly.
shonFlow is the finite-volume general CFD solver, described by the company as the first GPU-native Cartesian cut-cell CFD product in China, strong in moving machinery fluid simulation with moving boundaries, conjugate heat transfer that simulates fluid and solid together without separate meshes, user-defined functions and built-in visualization. It targets general scenarios competing with foreign general CFD.
shonTA combines a one-dimensional network with three-dimensional finite elements for system-level thermal management, used for three-dimensional thermal balance analysis of motors and reducers with conduction, contact, convection and radiation connections, coupled reduced-order models and virtual sensing, joint simulation with shonDy and shonMesh, built-in component templates and multi-objective optimization.
The three are method-complementary and scenario-layered: shonDy handles extreme free surfaces, shonFlow general flow and shonTA system-level thermal networks. The engineering meaning is that a complete electric drive analysis needs all three, using shonFlow for windage and cooling ducts, shonDy for churning and splashing and shonTA for whole-machine thermal balance.
Where Physical AI Sits
The 2026 recruitment of a physical AI algorithm engineer shows AI entering the roadmap. From public information the landing points are likely three.
The first is surrogate modeling. Particle simulation is computationally heavy, and fitting the mapping from operating parameters to temperature fields or losses with a neural network could replace part of the physical computation inside optimization loops, the same logic as Supreium's DPB and Ansys SimAI.
The second is reduced-order modeling. shonTA already mentions coupled reduced-order models and virtual sensing, compressing three-dimensional temperature fields into low-dimensional representations to support real-time or near-real-time system simulation, which is the technical basis for thermal digital twins.
The third is data-driven turbulence modeling or boundary condition calibration. Turbulence modeling and wall boundary treatment have long been weak points in particle methods, and data-driven improvement is a research frontier with insufficient industrial evidence.
The company has published no AI product name, metric or customer case. The physical AI engineer is a hiring signal rather than product evidence, and assessment should treat it as direction rather than current capability.
Technical Shortfalls
First, accuracy and conservation in particle methods. Least squares moving particle semi-implicit improves pressure oscillation over the classical method, but momentum and energy conservation remain less strict than in mesh methods and long integration accumulates error. For demanding cases such as laminar heat transfer coefficient calibration, mesh methods remain preferred.
Second, computational scale and cost. Each particle interacts with neighbors, so cost is approximately linear in particle count with a large constant, and GPU parallelism mitigates but does not eliminate the need for substantial compute on industrial models with hundreds of millions of particles.
Third, maturity of general CFD. shonFlow competes with Fluent, STAR-CCM+ and CFX, which hold decades of validation cases, turbulence model libraries and industry best practice. A new tool with architectural advantages in GPU-native design and Cartesian cut cells cannot close the validation gap quickly.
Fourth, material and working fluid databases. Thermal management simulation needs thermophysical data for refrigerants, lubricants and coolants, and the completeness of those data directly affects accuracy. Foreign tools ship large fluid libraries while domestic tools must build or license them.
Product and Commercial Model
Logic of the Product Matrix
The product line published on the company website divides into two groups. Solvers are shonDy, shonFlow and shonTA, providing fluid and thermal computation, and pre-processing is shonMesh for meshing. The 2026 recruitment material adds two design products to that list, shonEM for motor design and shonGear for drivetrain design, neither of which the website currently lists.
The design intent is to cover the fluid and thermal part of the electric drive development chain: meshing and pre-processing, then fluid and thermal analysis. The two design products mentioned in the recruitment material would extend that chain upstream to motor and drivetrain design. For customers that means completing most analysis inside one vendor's stack, reducing tool switching and data conversion.
A broad matrix does not mean every product is strong. shonFlow faces Fluent, STAR-CCM+ and CFX, and the two design products described in the recruitment material would face Motor-CAD, JMAG, Ansys Maxwell, Romax and KISSsoft, long-established specialists that new entrants struggle to displace. Their strategic value may lie more in customer binding, since combining the design products for motor design and shonTA for thermal analysis keeps data in one system and raises switching costs.
Revenue Scale and Business Model
The company discloses annual revenue in the tens of millions of renminbi, while registry financials for 2023 show 12.67 million renminbi. With a registry headcount of 34 insured employees and per-head cost of 400,000 to 600,000 renminbi, personnel cost is roughly 14 to 20 million; on the recruitment-platform range of 50 to 99 people the same calculation gives 20 to 60 million. Adding equipment, software and marketing brings total cost to perhaps 25 to 70 million depending on which headcount basis is used. If revenue sits between 12.67 million and 80 million, the company is likely near breakeven or slightly loss-making, consistent with having just closed a B round for expansion.
The model combines software licenses, engineering consulting and customized solutions. Licenses provide replicable revenue, consulting provides cash flow and customer insight, and customized projects bind leading customers. The combination is standard among Chinese CAE companies for a practical reason: pure licensing cannot cover costs early when customers are few and prices low, so services generate revenue while accumulating industry knowledge.
Customer Structure and Industry Mix
The distribution reflects strategic choice. New energy vehicles including Huawei, BYD, FAW, SAIC, Qingshan Industries and American Axle are the main battlefield, nuclear through CNNC is the second, and aerospace through AVIC plus marine through the Naval University of Engineering are technical validation scenarios.
The electric vehicle choice has clear reasons: thermal management and lubrication of electric drives are new problems, since conventional combustion vehicles have no high-speed motor churning loss, so foreign tools have limited accumulation, while China is the world's largest electric vehicle market with high customer density, fast iteration and willingness to trial new tools.
The nuclear choice is more interesting. Thermal hydraulic analysis in nuclear power demands extreme accuracy and validation because it touches safety analysis, and typically only certified codes are used. ShonCloud's entry there is probably not replacing safety analysis codes but design optimization, auxiliary analysis or component-level simulation. The commercial value is high price and long cycles, but the entry barrier is the highest.
International Footprint
The Karlsruhe office is a step toward internationalization with three likely purposes: recruiting European numerical talent, since Germany has strong particle method and CFD groups; reaching European automotive customers, since Germany is the industry center; and providing localization support for international sales.
Internationalization is a long-term necessity for Chinese CAE given the finite domestic ceiling, but it is expensive early with slow returns. A company with tens of millions in revenue opening a European research office shows willingness to invest for the long term.
Competitive Landscape
Rivals in Particle Methods
Commercial particle CFD tools are few. Internationally the representatives are Ansys Fluent with an SPH module that is not the core offering, Altair nanoFluidX based on SPH and focused on drivetrain lubrication and churning, and Simerics-MP with particle capabilities, alongside academic spinouts. Domestic particle work concentrates in universities such as Shanghai Jiao Tong, Dalian University of Technology and Huazhong University of Science and Technology, with few commercial products.
nanoFluidX is the most direct rival to shonDy, targeting the same churning, lubrication and splashing problems with the same meshless approach and customers including major automakers and transmission suppliers. The difference is ownership: nanoFluidX sits inside Altair, now part of Siemens, with global sales and brand backing, while shonDy is domestic with advantages in local service, price and deep collaboration with Chinese customers.
The Red Ocean of General CFD
shonFlow enters general CFD, a red ocean where Ansys Fluent, Siemens STAR-CCM+, Cadence Fidelity formerly NUMECA, COMSOL CFD and open-source OpenFOAM hold most share.
ShonCloud's differentiation is GPU-native architecture, Cartesian cut-cell meshing and moving boundaries. Cartesian cut cells generate meshes automatically with stable quality, suiting complex geometry and fast iteration, while boundary layer resolution trails body-fitted meshes and limits accuracy in high Reynolds number wall turbulence. GPU-native design offers cost advantages at equal compute.
Whether the differentiation holds depends on two conditions: whether target customers value fast iteration over ultimate accuracy, which electric vehicle thermal analysis does, and whether procurement policy provides access for domestic tools. Both currently hold, so shonFlow has room to exist without challenging the leaders.
Rivals in System-Level Thermal Management
shonTA faces Siemens Simcenter Amesim, Gamma Technologies GT-SUITE, Dassault Dymola and domestic peers, all with decades of one-dimensional system modeling and complete component libraries.
ShonTA's differentiation is coupling a one-dimensional network with three-dimensional finite elements. Conventional system simulation is mostly lumped parameter and cannot give internal component temperature distributions, while pure three-dimensional simulation is too slow. Combining both computes system-level flow and heat exchange in one dimension and key component temperature fields in three dimensions, balancing speed and accuracy. The logic suits electric drive thermal management because stator, winding and magnet temperature gradients require three-dimensional treatment while the cooling circuit can be one-dimensional.
The Common Position of Chinese CAE
Chinese CAE companies share two problems: validation accumulation and customer trust. Foreign tools have decades and tens of thousands of projects behind them, and results produced with them are accepted by customers, suppliers and regulators. A domestic tool with comparable algorithms needs extensive benchmarking to build trust.
ShonCloud's response is to build trust through specific scenarios and engineering consulting, using simulation to guide design inside customer projects, validating against measurements, and gradually expanding scope. The path is slow but solid, and it is the realistic choice for domestic CAE.
Financing and Capital Structure
Rounds and Use of Proceeds
Four rounds are recorded in public sources, with dates, amounts and investors as follows. The Pre-A round of December 23, 2020 was in the tens of millions of renminbi from Shunrong Capital and the Suzhou Xiangcheng fund, described as an angel round in the account of the local state-owned investment platform. The A round of January 28, 2022 was in the tens of millions of renminbi from Suzhou Guofa Venture Capital, Suzhou Asset Management and Xiangcheng state capital. The A+ round of December 28, 2023 came from Oriza Holdings with the amount undisclosed. The B round of March 19, 2025 was nearly 100 million renminbi from the Industrial Machine Tool Industry Investment Fund, Oriza Holdings, Shenzhen Hongtu Chongshi and Shenzhen Capital Group. An incubator brief describes the B round as a joint investment by the Industrial Machine Tool fund and Hongtu Chongshi, a subset of the four-party list. Valuations are not disclosed in any source.
Registry change records dated September 17, 2026 show the stake of Suzhou Shunrong Angel Fund IV decreasing, the stake of Shenzhen Hongtu Chongshi increasing, a change in the position of Wang Xi, departures of Li Gangqiang and Zhang Liming from principal membership and additions of Tong Zhou and Shi Chengpeng. The shareholder table still lists Shunrong Angel Fund IV at 2.6183 percent with a first-holding date of December 31, 2020, and Hongtu Chongshi first appears on March 11, 2025, the month the B round closed. The records therefore describe a shareholding adjustment and a board seat change, and Hongtu Chongshi was already an investor in the B round. A cumulative funding total of 1.40 hundred million renminbi appears only in an AI-generated summary on an investment portal and is treated here as an unverified estimate.
Stated use of proceeds is product iteration across the four official products and market expansion in electric vehicles, nuclear and aerospace. With tens of millions in revenue and a team in the tens, nearly 100 million renminbi covers roughly two to three years of operations and research.
Revenue and Cost Inference
With a self-reported revenue in the tens of millions and a registry headcount of 34, unit economics can be estimated loosely. At 50 million of revenue and 34 employees, revenue per head is about 1.47 million renminbi; at the 12.67 million of 2023 registry revenue and 34 employees, about 370,000. On the recruitment-platform range of 50 to 99 people, the same two revenue figures give roughly 500,000 to one million per head. Both spreads are wide, and they indicate the company remains in an investment phase with revenue not yet reflecting team capability.
On cost, a team including 11 PhDs and four foreign experts carries personnel cost well above an ordinary software company, and at a self-reported revenue in the tens of millions the company is likely at or slightly below breakeven.
Capital Efficiency and Future Needs
Seven years, four rounds and tens of millions of revenue with hundreds of customers is normal capital efficiency for Chinese CAE. Domestic companies generally need five to eight years to reach tens of millions in revenue, and ShonCloud's pace fits the pattern.
Future funding depends on two variables: how fast shonFlow penetrates general CFD, which requires more market investment, and internationalization spending on the German office and overseas customers. A listing in 2027 or 2028 would require revenue above 100 million renminbi with profitability.
2026 Tenders, Contracts and Credentials
Three procurement records dated 2026 separate by legal nature. In the Dongfeng Motor Corporation research institute procurement of September 9, 2026 for a 2026 shonDy v2.7 software lease, reference DF26Z-ZJ-FW-01253-01, the company appears as the successful supplier with a value of 275,000 renminbi. In the September 8, 2026 procurement by the China Institute for Radiation Protection and China Nuclear Environmental Protection for a model solving acceleration training service, the company appears as the third-ranked successful candidate with a quote of 651,000 renminbi, while the final award went to another party at 677,000 renminbi and was announced on September 20, 2026. In the October 20, 2026 procurement by Maiwang Environmental Technology for a dimensionless intelligent design model development service, the company appears as the second-ranked successful candidate with a quote of 963,000 renminbi, against 899,000 renminbi from the first-ranked candidate; the same procurement had failed to attract bids on September 30, 2026 and was re-organized. Only the first of the three is an award won by the company.
On credentials, the company was recognized as a national high-tech enterprise in 2025 under certificate number GR202532011194, published on December 19, 2025 and valid until December 19, 2028, and was listed among the second batch of Jiangsu specialized and sophisticated small and medium enterprises, published on December 5, 2025 and valid until December 5, 2028.
DISCREPANCIES
1. Revenue has only an informal source. The figure of tens of millions of renminbi appears in the company's 2026 campus recruitment material rather than a formal release or financing announcement, and neither the reference year nor the definition, whether contract value, recognized revenue or collections, is specified. This report retains the figure labeled as an informal channel. Source tier: company recruitment material, OFFICIAL but informal.
2. Two revenue statements of different order of magnitude. The recruitment material states annual revenue in the tens of millions of renminbi, while registry financials for 2023 record revenue of 12.67 million renminbi. The two may refer to different years or to different revenue definitions, and no public source reconciles them. This report presents both and does not adopt either as the revenue of record. Source tier: registry financial field (DB) and company statement (OFFICIAL) side by side.
3. Three customer counts. The company says hundreds of well-known enterprises, third-party material says about a hundred domestic and international customers, and conference material says close to a hundred. The figures may reflect different dates or definitions such as signed, active or paying customers. This report lists them side by side without adopting a single number. Source tier: multiple MEDIA sources with no unified OFFICIAL definition.
4. Three headcount bases. Registry records for 2025 list 34 insured employees, a recruitment platform lists 50 to 99 people, and an encyclopedia entry says more than 40. The three are not convertible into one another and none of them states a reference date. This report keeps them separate and uses no single headcount figure. Source tier: registry data (DB), recruitment platform (MEDIA) and encyclopedia (MEDIA).
5. Three versions of the talent roster. The July 2025 district media release lists 11 full-time PhDs, 2 national-level talents, 8 provincial or municipal talents and 4 foreign experts with over 70 percent of staff holding a master's degree or above. The 2026 recruitment material keeps the 70 percent share and the 4 foreign experts but lists 1 national-level talent and 6 provincial or municipal talents. An AI-generated summary on an investment portal lists 6 full-time PhDs including 2 foreign experts and 1 national-level talent. This report treats the release and the recruitment material as the primary accounts and the portal summary as unverified. Source tier: MEDIA, OFFICIAL and UNVERIFIED in parallel.
6. Naming difference in the particle method. One source describes shonDy as based on moving particle semi-implicit, another as least squares moving particle semi-implicit. The latter is an improved variant applying least squares discretization for better accuracy and stability, so the two are consistent but technically different. This report uses the least squares formulation and explains its relation to the classical method. Source tier: company product material (OFFICIAL) over third-party summary.
7. Exclusivity of the first GPU-native Cartesian cut-cell CFD claim. The company describes shonFlow as the first GPU-native Cartesian cut-cell CFD product in China. The first claim depends on how the intersection of GPU-native, Cartesian cut-cell and CFD is defined and cannot be independently verified. This report does not use the exclusivity claim. Source tier: company marketing, UNVERIFIED.
8. Funding rounds, investors and the four-round timeline. The most complete account records a Pre-A round on December 23, 2020 in the tens of millions from Shunrong Capital and the Suzhou Xiangcheng fund, an A round on January 28, 2022 in the tens of millions from Suzhou Guofa Venture Capital, Suzhou Asset Management and Xiangcheng state capital, an A+ round on December 28, 2023 from Oriza Holdings with the amount undisclosed, and a B round on March 19, 2025 of nearly 100 million from the Industrial Machine Tool Industry Investment Fund, Oriza Holdings, Shenzhen Hongtu Chongshi and Shenzhen Capital Group. A separate incubator brief describes the B round as a joint investment by the Industrial Machine Tool fund and Hongtu Chongshi, a subset. Valuation figures are not disclosed in any source. This report lists the four rounds with dates and investors and states which source each entry comes from. Source tier: registry data (DB) as the most complete account, with a MEDIA account of the B round as a subset.
9. A cumulative funding total that is an estimate. The figure of 1.40 hundred million renminbi in cumulative funding appears only in an AI-generated summary on an investment portal, which marks it as an inferred summary rather than a company disclosure. This report records it as an estimate that has not been confirmed against a primary source. Source tier: AI-generated portal summary, UNVERIFIED.
10. Shareholding change versus exit. Registry change records dated September 17, 2026 show the stake of Suzhou Shunrong Angel Fund IV decreasing, the stake of Shenzhen Hongtu Chongshi increasing, a change in the position of Wang Xi, departures of Li Gangqiang and Zhang Liming from principal membership and additions of Tong Zhou and Shi Chengpeng. The shareholder table still lists Shunrong Angel Fund IV at 2.6183 percent with a first-holding date of December 31, 2020, and Hongtu Chongshi first appears on March 11, 2025, the month the B round closed. The records describe a stake adjustment and a board seat change, and Hongtu Chongshi was already an investor in the B round. Source tier: registry change record (DB).
11. Three 2026 procurement records of three different natures. In the Dongfeng Motor Corporation research institute procurement of September 9, 2026, the company is the successful supplier for a shonDy v2.7 software lease with a contract value of 275,000 renminbi. In the September 8, 2026 procurement by the China Institute for Radiation Protection and China Nuclear Environmental Protection, the company is the third-ranked successful candidate for a model solving acceleration training service, quoting 651,000 renminbi against a winning price of 677,000 renminbi. In the October 20, 2026 procurement by Maiwang Environmental Technology, the company is the second-ranked successful candidate for a dimensionless intelligent design model development service, quoting 963,000 renminbi against 899,000 renminbi from the first-ranked candidate. Only the first record is an award won by the company. This report keeps the three natures distinct. Source tier: procurement announcements (OFFICIAL).
12. Registered entity versus the cities named in public materials. The registered address is in Xiangcheng district, Suzhou, and some third-party materials describe the company as Shenzhen-based, while its subsidiaries are located in Shenzhen and Beijing. The operating entity is therefore stated as Suzhou with Shenzhen and Beijing as subsidiary locations. Source tier: registry record (DB) and third-party profiles (MEDIA).
13. Product line count. The company website lists four products, shonDy, shonTA, shonFlow and shonMesh, while the 2026 recruitment material names six, adding shonEM and shonGear. This report describes the four official products and notes the two additional names as a recruitment-material version. Source tier: company website (OFFICIAL) and recruitment material (OFFICIAL).
14. Relationship between the physical AI role and AI products. The 2026 campus recruitment includes a physical AI algorithm engineer, indicating hiring in that direction but not an existing product or customer case. Public information contains no AI product name, metric or application. This report treats AI as direction rather than capability. Source tier: recruitment information (OFFICIAL) contrasted with product information.
15. Nature of the nuclear customer relationship. The customer list includes CNNC without specifying whether the work is safety analysis, design optimization or component-level simulation. Safety analysis codes in nuclear power require certification by the national nuclear safety regulator, which new domestic tools cannot enter quickly. This report distinguishes possible scope in customer analysis without overreaching. Source tier: company customer list (OFFICIAL) with business detail absent.
Risks and Diligence Gaps
First, revenue scale and profitability are unconfirmed. The tens of millions figure comes from recruitment material rather than an audited source, and with 11 PhDs and four foreign experts on staff the company may be loss-making. This is the foundational gap.
Second, the market ceiling in the niche. Particle methods lead in churning, splashing and wave impact, which are niches within niches. The number of customers doing drivetrain lubrication simulation worldwide is limited, and that alone cannot support scale, making expansion into general CFD and system thermal management necessary but crowded.
Third, the validation gap in general CFD. Competing with Fluent and STAR-CCM+ requires extensive validation cases, turbulence libraries and best practice accumulated over time and projects, which cannot be closed quickly.
Fourth, cost pressure from the talent structure. Eleven full-time PhDs, four foreign experts and two national talents carry high compensation, and at tens of millions of revenue personnel cost may exceed half of revenue and constrain research and marketing spending.
Fifth, customer concentration and industry cycles. Customers concentrate in electric vehicles, exposed to automotive price competition and research budgets, and in nuclear, exposed to project approval pacing.
Sixth, the input-output balance of internationalization. The German office and overseas development require sustained investment while acceptance of Chinese CAE abroad is low on brand, validation and service network, making the near term a net investment.
Seventh, the AI validation gap. The physical AI engineer is a hiring signal without products or cases, and if customers expect AI features before the product is ready it would weaken the sales narrative.
Composite Assessment
Technology scores 7.5. Least squares moving particle semi-implicit offers genuine advantage on free surfaces and moving boundaries, GPU and multi-GPU distributed computing reach engineering usability, the four official products cover the fluid and thermal part of the electric drive development chain, and the one-dimensional plus three-dimensional thermal coupling is sound. Deductions for conservation and accuracy limits in particle methods, insufficient validation in general CFD, and unproductized AI.
Business scores 7.0. Revenue in the tens of millions against 12.67 million in 2023 registry data, customers including Huawei, BYD, FAW, SAIC and CNNC, a precise and painful entry scenario in oil churning, a pragmatic mix of software, consulting and customization, and three 2026 procurement records that include one award won as the successful supplier. Deductions for small revenue, a low niche ceiling and intense competition in general CFD.
Finance scores 6.5. Four rounds with a B round of nearly 100 million shows normal capital efficiency, and the customer and industry mix is sensible. Deductions for unreconciled revenue statements, a position near breakeven and a high personnel cost ratio.
Product scores 7.5. The official four-product line is complete for fluid and thermal analysis, shonDy is differentiated in churning, shonTA's one-dimensional and three-dimensional coupling has engineering value, and shonFlow's GPU-native architecture offers cost advantage. Deductions for the pressure the two design products named in the recruitment material would face from specialists, a working fluid database that must be built in house, and uneven product maturity.
Market scores 7.0. China leads the world in electric vehicle scale, electric drive thermal management is a new problem, procurement policy provides access for domestic tools, and nuclear plus aerospace offer high-value scenarios. Deductions for a finite niche, foreign strength in general CFD and dedicated design tools, and cyclical customer industries.
Overall 7.1. This is the most precisely positioned technology choice among domestic CAE companies in this series: it entered churning and splashing, where conventional tools struggle, built credibility on a concrete engineering pain point, and then expanded into general and system-level simulation. Its weakness is scale, with a self-reported revenue in the tens of millions against 12.67 million in 2023 registry data, a low niche ceiling and an insufficient validation base in general CFD. The judgment on physical AI is that its physical core in particle free surfaces and one-dimensional plus three-dimensional thermal coupling is real and hard to replace while AI is a direction being laid out rather than existing capability, and that growth from niche specialist to platform vendor depends on shonFlow establishing itself in the general market and on AI surrogates genuinely lowering the cost of particle simulation.
References
- ShonCloud 2026 campus recruitment material: six years since founding, completed a B round of nearly 100 million renminbi, annual revenue in the tens of millions; national high-tech enterprise, Suzhou key laboratory for industrial multiphysics simulation software, member of the fifth China automotive hidden unicorn list; 11 full-time PhDs, two national-level talents, eight provincial and municipal talents, four foreign experts; headquarters in Suzhou with subsidiaries in Shenzhen and Beijing and a German research center; customers including Huawei, FAW, CNNC and BYD among hundreds of enterprises; six flagship products shonDy, shonTA, shonFlow, shonMesh, shonEM and shonGear; positions open for fluid numerical algorithm engineer, high performance parallel algorithm engineer, physical AI algorithm engineer, electromagnetic numerical algorithm engineer, mesh algorithm engineer and graphical interaction software engineer
- Product descriptions in the same material: shonDy as moving particle semi-implicit software used for bearing oil injection lubrication simulation; shonTA as one-dimensional and three-dimensional multiphysics system simulation used for turbine blade coupled computation including external flow import, internal pipe flow and three-dimensional temperature fields; shonFlow as finite-volume general fluid simulation used for transient moving boundary computation of an automotive torque converter; shonMesh as meshing software supporting Python scripts for automatic geometry import and mesh generation; shonEM as motor design software supporting parametric modeling, winding design, single point simulation and efficiency maps; shonGear as drivetrain design software for speed and torque, power loss and tooth contact analysis
- Third-party company registry, innohere: Suzhou ShonCloud Engineering Software Co., Ltd., established June 26, 2019, legal representative Wang Xi, address in Xiangcheng district Suzhou, 50 to 99 employees, four funding rounds, two patents; main business in shonDy, shonTA and shonFlow fluid multiphysics software and technical services
- Conference and media material: founded June 2019 and focused on next-generation fluid multiphysics engineering simulation software; customers in new energy vehicles, nuclear, energy storage, pumps and valves and aerospace; more than 70 percent of research staff holding master's degrees or above; Suzhou headquarters research center with research branches in Beijing, Shenzhen and Karlsruhe; multiple funding rounds completed, serving hundreds of leading enterprises including Huawei, SAIC, BYD and CNNC
- Product technical descriptions in the same material: shonDy based on least squares moving particle semi-implicit, meshless and strong in splashing and fluid-structure interaction, using GPU acceleration and multi-GPU and multi-CPU distributed computing with integrated oil shedding and fluid breakup; shonTA as a one-dimensional network plus three-dimensional finite element system multiphysics product focused on electric vehicle thermal management with coupled reduced-order models and virtual sensing, joint simulation with shonDy and shonMesh, component templates and multi-objective optimization; shonFlow as the first GPU-native Cartesian cut-cell CFD product in China, strong in moving machinery, with strengthened moving boundaries, a built-in physical model library, user-defined functions and built-in visualization
- Third-party vendor profile, updated August 17, 2026: founded in 2019 by Wang Xi, headquartered in Suzhou with subsidiaries in Shenzhen and Beijing and an office in Karlsruhe; about a hundred domestic and international customers in new energy vehicles, energy and power engineering and manufacturing; shonDy as a high performance CFD solver; shonTA based on thermal network and finite element methods supporting conduction, contact, convection and radiation connections and mainly used for three-dimensional thermal balance of motors and reducers; shonFlow based on finite volume, supporting steady and transient analysis, rotating machinery tools and conjugate heat transfer
- Recruitment information from Liepin: the company lists customers including Huawei, CNNC, AVIC, American Axle, Qingshan Industries, FAW and SAIC plus universities including the Naval University of Engineering, Chongqing University of Technology, Beijing Institute of Technology and Shenzhen University; honors include high-tech enterprise, specialized and sophisticated SME, Jiangsu dual-innovation talent enterprise and Jiangsu technology-based SME; headquarters and main research in Suzhou with research offices in Beijing and Karlsruhe
- Technical background from public sources: moving particle semi-implicit approximates field variables with kernel functions and needs no mesh, representing free surfaces by particle presence and moving boundaries by moving boundary particles; the least squares variant uses least squares spatial discretization to improve accuracy and stability and alleviate pressure oscillation; costs include harder accuracy and conservation control, heavy computation and complex boundary treatment
- Engineering background from public sources: electric drive systems use oil cooling and lubrication, with oil agitated and thrown by gears to form films and mist, and churning loss representing a meaningful share of drivetrain loss at high speed affecting range; optimization covers oil level, gear geometry, housing guides and oil passages, all requiring knowledge of oil location, velocity and film thickness
- Competitor information from public sources: particle commercial tools include Altair nanoFluidX based on SPH and focused on drivetrain lubrication and churning, the Ansys Fluent SPH module and Simerics-MP; general CFD is dominated by Ansys Fluent, Siemens STAR-CCM+, Cadence Fidelity, COMSOL CFD and open-source OpenFOAM; system thermal competitors include Siemens Simcenter Amesim, Gamma Technologies GT-SUITE and Dassault Dymola; motor and drivetrain design competitors include Motor-CAD, JMAG, Ansys Maxwell, Romax and KISSsoft
- Capital and credential information: four funding rounds completed with the latest B round at nearly 100 million renminbi; national high-tech enterprise, Suzhou key laboratory for industrial multiphysics simulation software, member of the fifth China automotive hidden unicorn list, Jiangsu dual-innovation talent enterprise and Jiangsu technology-based SME
- Industry background from public sources: nuclear thermal hydraulic safety analysis uses codes certified by the national nuclear safety regulator such as RELAP5 series and certified CFD versions, and new tools require long validation and certification to enter safety analysis; Chinese CAE companies generally need five to eight years to reach tens of millions in revenue, with validation accumulation and customer trust as the main barriers
- Company official website, product and about pages: the published product line comprises four items, shonDy meshless CFD, shonTA multiscale thermal analysis, shonFlow finite-volume CFD and shonMesh pre-processing; the site does not list shonEM or shonGear
- Venture capital database profile: legal name Suzhou ShonCloud Engineering Software Co., Ltd., incorporation date June 26, 2019, registered address at 26-27/F Block A, Huo Li Business Plaza, 185 Jumao Street, Yuanhe Subdistrict, Xiangcheng District, Suzhou; legal representative Wang Xi; four funding rounds
- Business registry change record dated September 17, 2026: stake of Suzhou Shunrong Angel Fund IV decreased, stake of Shenzhen Hongtu Chongshi increased, change in the position of Wang Xi, departures of Li Gangqiang and Zhang Liming from principal membership, additions of Tong Zhou and Shi Chengpeng; the shareholder table lists Shunrong Angel Fund IV at 2.6183 percent with a first-holding date of December 31, 2020 and Hongtu Chongshi with a first-holding date of March 11, 2025
- Business registry financial field: revenue of 12.67 million renminbi for 2023; insured employees 34 in the 2025 annual report
- Dongfeng procurement announcement, September 9, 2026: 2026 shonDy v2.7 software lease, reference DF26Z-ZJ-FW-01253-01, purchaser Dongfeng Motor Corporation research institute, successful supplier ShonCloud, 275,000 renminbi
- Procurement candidate announcement, September 8, 2026: model solving acceleration training service, purchaser China Institute for Radiation Protection and China Nuclear Environmental Protection, third-ranked successful candidate ShonCloud at 651,000 renminbi; final award 677,000 renminbi announced September 20, 2026
- Procurement candidate announcement, October 20, 2026: dimensionless intelligent design model development service, purchaser Maiwang Environmental Technology under China Energy Conservation, second-ranked successful candidate ShonCloud at 963,000 renminbi against 899,000 renminbi from the first-ranked candidate; the procurement failed to attract bids on September 30, 2026 and was re-organized
- Jiangsu provincial announcement, December 19, 2025: national high-tech enterprise certificate number GR202532011194, valid until December 19, 2028
- Jiangsu provincial announcement, December 5, 2025: second batch of specialized and sophisticated small and medium enterprises, valid until December 5, 2028
- District media release, July 7, 2025: over 70 percent of research staff holding master's degrees or above, 11 full-time PhDs, 2 national-level talents, 8 provincial and municipal talents, 4 foreign experts; the same text states a completed A+ round of nearly 100 million renminbi and annual revenue in the tens of millions of renminbi
- Incubator brief: the B round described as a joint investment by the Industrial Machine Tool Industry Investment Fund and Shenzhen Hongtu Chongshi
- Investment portal AI-generated summary: cumulative funding of 1.40 hundred million renminbi marked as an AI summary, and an encyclopedia entry listing more than 40 employees
This report is compiled from public sources for informational purposes only and is not investment advice. Data is current as of 2 November 2026. The company is private and revenue figures come from recruitment material and registry fields rather than audited accounts; customer counts and market share claims are company-reported without third-party verification, and related judgments are labeled as uncertain.
Related papers
Related cases
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
- 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
- Shichuang Technology Deep Dive: SupreCAST Cloud-Native Material-Processing Simulation and Industrial Agents Advance the CAE+AI Strategy
- 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.