Yunqi Capital Leads Angel Round in Youshu Quantum, as Shanghai Jiao Tong University Team Explores Quantum Scientific Computing | Yunqi Partners
Providing new algorithmic pathways for solving scientific and engineering problems in quantum computing

Fluid simulation, options pricing, electromagnetic computation — these science and engineering problems all ultimately come down to solving equations. The objects of study are getting more complex, the equations are getting harder to compute, and classical compute is starting to look insufficient. Quantum computing can fill exactly this gap — provided someone first turns real-world problems into a quantum-solvable form.
That's precisely what the "Schrödingerization" algorithm framework, originally developed by a Shanghai Jiao Tong University team, does: through mathematical transformation, it converts differential equations into a form suited for quantum simulation. UnitaryLab — wait, that's the platform name — Unitary Quantum has built around it the agent-driven quantum scientific computing platform UnitaryLab and the desktop-level heterogeneous computing base UnitarySpark, and has run experiments on real hardware with QuantumCTek, cutting gate counts by more than 90%. Scenario partners including Bank of Shanghai have already signed on.
The company recently announced the completion of its angel round, led by the Yunqi–SJTU AI Angel Fund. In this edition of Yunqi Partners, we share the latest from Unitary Quantum.

Yunqi Capital's investment view:
We continue to watch how quantum computing lands across industries. In that process, algorithms are the core link that takes quantum computing toward general-purpose viability. Unitary's team builds on the algorithm-level Schrödingerization approach to extend quantum computing capability into general-purpose computing, which has the potential to become an important solution for bringing quantum computing to market.

Shanghai Unitary Quantum Technology Co., Ltd. (hereinafter "Unitary Quantum") recently announced the completion of its angel round. New investors in this round are Yunqi Capital and Jiaming Haochun, with existing shareholders Hanyuan Asset Management, Xiaomiao Langcheng, and INS Ventures continuing to increase their positions. Multi-Dimensional Capital has long served as exclusive financial advisor. The funds will be mainly used for original quantum algorithm R&D, iteration of the UnitaryLab quantum scientific computing platform and the UnitarySpark heterogeneous computing product, optimization of AI and agent-driven capabilities, quantum hardware adaptation, and expansion into research and industrial markets.
Less than three months after closing its previous round, Unitary Quantum has successively completed a software platform upgrade, launched a desktop-level heterogeneous computing product, carried out engineering scenario validation involving a domestically built general-purpose quantum computer, and integrated its self-developed algorithm tools into a third-party research agent platform.
Original "Schrödingerization" Algorithm
Exploring a Quantum Path for Scientific Computing
Whether it's fluid dynamics simulation, financial options pricing and credit risk, electromagnetic simulation and power flow computation, seismic wave solving in geological exploration, or wave and particle motion in complex physical systems, scientists and engineers need to turn real-world problems into mathematical equations and then solve them with computational software. As the objects of study grow more complex and the requirements for computational scale and precision keep rising, solution efficiency itself has become a core constraint on solving complex research and engineering design problems. Quantum computing is meant to solve the problems that AI and classical compute struggle with or handle poorly.
Unitary Quantum's technical foundation is the "Schrödingerization" algorithm framework originally developed by a research team at Shanghai Jiao Tong University. In 2024, related research was published in high-level international journals including Physical Review Letters and PNAS, and was selected as a representative achievement in the National Natural Science Foundation of China's 2024 annual report and in the compilation of outstanding achievements of the 14th Five-Year Plan. The method uses mathematical transformation to convert linear ordinary and partial differential equations — as well as some important nonlinear differential equations — into a form suited for quantum simulation, offering a new algorithmic path for quantum computing to solve science and engineering problems and filling a gap in the field.
Rather than developing a single quantum algorithm for a specific scenario, Unitary Quantum aims to develop this mathematical method into an algorithm and software system applicable to different scientific computing tasks, connecting scientific problem modeling, algorithm design, simulation validation, and execution on quantum computers.
AI Agent-Driven UnitaryLab
Making Complex Algorithms Easier to Use
In May 2026, Unitary Quantum released the agent-driven quantum scientific computing platform UnitaryLab 2.0, and in September launched the public beta of UnitaryLab 2.5, further integrating self-developed algorithm tools, agent task orchestration, AI quantum circuit editing, quantum simulation, GPU acceleration, and real quantum hardware invocation.
Traditional scientific computing typically involves multiple steps — mathematical modeling, code writing, parameter configuration, computational task execution, and results analysis — placing high demands on researchers' software development skills. By introducing AI agents, UnitaryLab lets researchers describe scientific problems in natural language, with the agent assisting in algorithm selection, parameter configuration, AI quantum circuit editing, simulation execution, quantum computer execution, and results analysis, reducing repetitive operations in complex computational tasks.
This capability has already been demonstrated in real research workflows. On August 26, NVIDIA's official technical blog introduced a local quantum scientific computing workflow built on DGX Spark, NVIDIA NIM, and the UnitaryLab Agent. Taking a two-dimensional heat conduction problem as an example, researchers specify boundary conditions and computational requirements in natural language, and the agent assists in completing the relevant computational tasks. In that case, roughly 12 manual steps were reduced to about 3 natural-language interactions.
Beyond its own platform, Unitary Quantum is also integrating its algorithm tools into third-party research agent ecosystems. On October 22, the company's self-developed quantum-skills quantum computing skill components went live on Shanghai Artificial Intelligence Laboratory's "Intern·Duan Yan" scientific discovery platform, covering 12 technical domains and 51 algorithm and workflow guides, with UnitaryLab as the default simulator to help research agents retrieve, understand, and use relevant quantum algorithms.
UnitarySpark Connects Heterogeneous Compute
Advancing Real Quantum Hardware Applications
In September 2026, the company launched UnitarySpark, a desktop-level heterogeneous computing hardware base that integrates AI agents, quantum simulation, and GPU computing power into a local computing environment, with interfaces reserved for calling external quantum computers. The product targets R&D or teaching teams at universities, research institutions, quantum software and hardware companies, and scenario-partner enterprises that need localized deployment. It meets the desktop-level, localized, hundred-thousand-yuan-budget hardware demand created by the emerging quantum computing R&D market, providing hardware support for algorithm development, research validation, joint R&D, and quantum computing talent training.
On the hardware ecosystem side, Unitary Quantum has signed strategic cooperation agreements with QuantumCTek, Bose Quantum, and SinoQuantum, carrying out close collaboration on software-hardware adaptation, quantum real-machine connectivity, and application validation, covering different technology routes including superconducting, photonic, and neutral atom.
In July 2026, Unitary Quantum and QuantumCTek completed experiments on heat and convection equations with 10 qubits on the latter's quantum computer, corresponding to 16 spatial grid points. The representative control group covered 5 time points for each of the two equation types, completing 10 circuit executions and 40,960 samplings in total, connecting the full pipeline of circuit optimization, native gate compilation, chip mapping, real-machine execution, and energy recovery. Compared with the baseline of directly compiling the original circuits for these instances, total gate count was reduced by more than 90%. Multi-time-step experimental controls have now been established, and the two sides will continue to improve real-machine precision and evaluate the stability of repeated runs.
In August 2026, Unitary Quantum and Bose Quantum carried out hybrid computing validation in a computational fluid dynamics (CFD) scenario. In a two-dimensional flow-around-a-cylinder task, the two sides used Bose's special-purpose quantum computer to optimize key collocation points, combined with a physics-informed neural network (PINN) to complete flow field prediction. In this experiment, quantum hardware handled specific optimization tasks while the classical neural network handled subsequent training and prediction, preliminarily validating the feasibility of special-purpose quantum compute participating in engineering computation workflows.
At the same time, Unitary Quantum continues to deepen technical cooperation with GPU companies including NVIDIA and Biren Technology, building a quantum computing intelligent compute ecosystem compatible with diverse compute architectures on the back of its self-developed UnitaryLab full-amplitude quantum simulator backend. The company has achieved preliminary adaptation results on mainstream GPUs and domestic GPU platforms, gradually connecting the full technical chain from underlying GPU compute and quantum simulators to quantum algorithm libraries and intelligent agents.
Through UnitarySpark and hardware cooperation across different technology routes, Unitary Quantum is gradually building a unified execution system connecting original quantum algorithms, classical heterogeneous compute, and quantum hardware. This system enables the company to fully leverage existing GPUs, quantum simulators, and special-purpose/general-purpose quantum computers for product R&D and application validation, reducing dependence on the maturity timeline of future large-scale fault-tolerant quantum computers and accelerating the engineering application and commercialization of quantum computing technology.
From Research Tool to Industrial Application
Exploring a Commercialization Path for the Scientific Computing Platform
In September 2026, at its product launch event, Unitary Quantum held scenario cooperation signing ceremonies with several enterprises and institutions, including Bank of Shanghai, the Hefei Quantum Technology Innovation Center for Meteorology, Changchun Jinsai Pharmaceutical, Shanghai Normal University, Chongqing University of Posts and Telecommunications, and Zanqi Technology, covering fields such as finance, meteorology, biomedicine, artificial intelligence, and quantum education and teaching.
The company is mainly exploring business with three types of customers: universities and research institutions, to which it offers algorithm platforms, science education platforms, research toolchains, and localized deployment solutions; quantum hardware companies, with which it explores software adaptation, algorithm services, and agent toolchain cooperation; and industry customers in finance, simulation, automotive and new energy, life sciences, meteorology and aerospace, and communications and electromagnetics, where it seeks application opportunities with clear computational needs through joint R&D and scenario validation. The company has already moved from productization toward commercialization, achieving a zero-to-one breakthrough in commercial cooperation across multiple fields.
The above content is excerpted from the WeChat account "UnitaryLab."


