For stock investors, the Gold Qilin analyst research reports offer authoritative, professional, timely and comprehensive insights to help uncover potential thematic opportunities.
Two former classmates from Fudan University are pushing a company less than a year old onto the global quantum computing stage.
Taichi Quantum, a neutral-atom quantum computing firm, has completed a new funding round led by existing shareholder Huakong, with participation from Jinqiu Fund, Chengdu Gaoxin Vision Fund, Yunbai Capital, Lingang Frontier Investment, Hongruida Investment, Ningbo Huihao and others, while existing shareholders Hengxu Capital, Boyuan Capital, Yarui Capital and Gaomai Capital continued to increase their stakes. Since its founding in January 2026, Taichi Quantum has raised a cumulative total of more than 1 billion yuan, less than four months after its previous 300 million yuan Pre-A round.
In a quantum computing industry that is highly cutting-edge and still lacks clear valuation benchmarks, Taichi Quantum is one of the domestic neutral-atom companies with the strongest fundraising appeal. It is the first startup in this track in China to publicly announce cumulative financing of over 1 billion yuan. After this round, the company's valuation has reached unicorn level.
Existing shareholders are best positioned to judge whether a company is viable. The lead investor in this round, Huakong, already appeared on the shareholder list in the June Pre-A round. Founded in 2007 and originating from Tsinghua University, this fund has recently been very active in frontier technology fields such as AI, quantum computing and embodied intelligence. This round carries a clear industrial flavor, with investor backgrounds spanning AI and computing power, electronic information, automotive and new energy, advanced manufacturing, and healthcare. This reflects that Taichi Quantum has begun to consider financing together with future computing coordination and application scenarios.
Beyond its efficient financing pace, Taichi Quantum has assembled a team of nearly 100 people over the past few months, completed the integration of its first prototype machine in about six months, and advanced the development of real-time control systems and a quantum compiler. In a frontier industry accustomed to discussing long-term visions, the two post-90s Fudan founders are trying to lead by example, accumulating bargaining chips for the next stage of competition through organizational and engineering progress.
From an arms race in physical qubits to a new consensus on fault tolerance
Over the past few years, the global quantum consensus has loosened, and neutral atoms have rapidly gained prominence due to potential advantages such as scalability, with companies including Atom Computing, QuEra and Pasqal successively advancing system development. In March this year, Microsoft (NASDAQ: MSFT), which had long focused on the superconducting route, also announced a new neutral-atom research team, pursuing both routes in parallel.
Taichi Quantum is one of these new neutral-atom forces. Founded in January 2026, founder and CEO Fang Zhenghao has nearly 15 years of experience in frontier technology investment and industry practice, having invested in and incubated more than 100 technology companies. Founder and CTO Liu Hongbin previously served as chief architect of quantum at Microsoft (NASDAQ: MSFT), with full-chain R&D experience ranging from quantum algorithms and quantum error correction to system architecture and full-machine engineering.
But regardless of the route, a more fundamental choice must be faced. Over the past few years, quantum computing has experienced an arms race in the number of physical qubits, from dozens to hundreds, to thousands and even larger arrays. Afterwards, the global quantum industry began to form a new consensus: for quantum computing to truly enter enterprise-level applications, the difficulty has long ceased to be merely "how many physical qubits can be made," because physical qubits make errors, and as operations accumulate, tiny errors can render results meaningless.
The focus of quantum computing competition has gradually shifted from who has more physical qubits to who can first build a truly operational logical qubit system. In September this year, Microsoft (NASDAQ: MSFT) quantum head and computational physicist Matthias Troyer, Microsoft (NASDAQ: MSFT) quantum hardware and systems engineering head Chetan Nayak, and 2025 Nobel Prize in Physics laureate John Martinis of Qolab further proposed a definition of "scalable logical qubits": the ability to maintain long-time computation through repeated error correction, support fault-tolerant universal operations, feature low-latency real-time decoding and feedback, and scale to hundreds or thousands as applications require.
This set of definitions examines reliability, scale, computing capability and performance together. Demonstrating a logical qubit is only the starting point; how long it can run, what operations it supports, and how many physical resources it requires also determine practical value. Specific applications have different requirements for error rates and computational depth, and commercialization thresholds cannot be summarized by a single number.
As a result, the arms race has not ended, but has extended from quantity to system capability. Overseas companies still diverge on technology routes, but fault-tolerant quantum computers are recognized as the next battleground in the global quantum computing race. Taichi Quantum's choice of the ytterbium atom route from the outset is precisely a pragmatic backward derivation under this framework: first determine the fault-tolerance goal, then consider error correction schemes, control architecture and physical systems. From its founding, the company concentrated resources on universal fault-tolerant computing.
In Fang Zhenghao's view, the most important decision at Taichi Quantum's inception was not how many physical qubits to build first, but to determine the future fault-tolerant computing architecture, and to organize technology routes, talent and capital investment around that endgame. Compared with the traditional path relying on university laboratories, Taichi Quantum has organized interdisciplinary R&D in corporate form from the start, putting engineering and productization on the same timetable.
In Fang Zhenghao's view, the real barrier to deep-tech entrepreneurship comes not only from the technology route itself, but also from continuously gathering world-class talent, long-term capital and industrial resources around the correct technology endgame, and enabling these elements to reinforce one another.
Less than a year after its founding, Taichi Quantum has demonstrated highly efficient talent organization. According to disclosures, the company has recruited a team of nearly 100 people, with members from universities such as MIT, Cambridge, the University of Pennsylvania, LMU Munich, Tsinghua and Peking University, research institutions such as JILA and CQT, and companies such as Microsoft (NASDAQ: MSFT) and Zurich Instruments, covering physics, optics, control, software and algorithms.
Notably, it is said that Taichi Quantum has attracted foreign scientific research and engineering personnel, with several members choosing to join Taichi Quantum even after receiving offers from companies such as QuEra and Oratomic. Behind this talent appeal is a full-stack self-development plan aimed directly at global giants: from physical hardware and measurement and control to compilation, error correction and algorithm implementation, the company hopes to master key links.
Neutral-atom quantum computing systems involve multiple complex subsystems including vacuum, lasers, optics, atomic manipulation, imaging, measurement and control, real-time control and software. Full-stack means no single point can have weaknesses, which can ultimately form higher barriers, improve system R&D efficiency, and create conditions for long-term performance optimization, but the difficulty of system implementation also increases exponentially.
At present, full-stack R&D is a common choice for leading overseas companies, with Microsoft (NASDAQ: MSFT) and other top enterprises providing important references. Behind Willow's error correction progress, chip design and manufacturing, quantum operations, calibration and real-time decoding need to be optimized together, ultimately tested by logical error rates. Whether different links can share problems and iterate quickly directly affects overall performance.
However, full-stack self-development is relatively rare in China's quantum industry, and Taichi Quantum may be one of the few neutral-atom companies in China that self-develops all core technologies across the stack. CTO Liu Hongbin's philosophy is to implement this kind of system coordination within the company, and full-stack self-development is expected to achieve the highest R&D efficiency and overall performance.
Supply chain maturity is also a major challenge. Although industries such as precision manufacturing and optoelectronics have provided a good foundation, some key components and system interfaces for fault-tolerant quantum computing still need continuous optimization. Through full-stack R&D and system integration, Taichi Quantum enables physics, control and software teams to collaborate around a unified architecture. Therefore, full-stack is both a technical choice and an organizational choice. It requires simultaneously recruiting talent in multiple scarce directions and turning interdisciplinary communication into a daily engineering process, exchanging higher upfront investment for R&D autonomy and iteration efficiency.
Betting on full-stack, beginning to deliver results
This full-stack plan has already produced concrete engineering results. Recently, Taichi Quantum disclosed three advances in full-machine integration, real-time control and a quantum compiler, corresponding respectively to the physical platform, control feedback and software capabilities.
On one hand is efficient engineering efficiency; on the other hand, it is inseparable from the CTO's mature experience. Liu Hongbin previously led teams to achieve demonstrations and engineering implementation of fault-tolerant quantum computing systems multiple times, with system-level design, integration and delivery experience for fault-tolerant computing.
Taichi Quantum took about six months to achieve full-machine integration of its first prototype, connecting modules such as vacuum, lasers, optics, atomic manipulation, imaging, and measurement and control into a unified system, entering the joint debugging and optimization stage.
In terms of the control system, Taichi Quantum has advanced system real-time control latency to the order of 10 microseconds. Fault-tolerant systems need to receive measurement results in a timely manner, judge errors and decide subsequent operations, and classical computing and quantum hardware must work closely together. Cutting-edge players such as Quantinuum and Atom Computing are strengthening real-time control capabilities for error correction, and NVIDIA is also promoting microsecond-level real-time connections between GPUs and quantum processors. Taichi Quantum's advancement of real-time control latency to the order of 10 microseconds means it has initially opened the closed loop from camera input and data processing to atomic rearrangement, and is advancing coordination among FPGAs, GPUs and quantum hardware.
At the software level, Taichi Quantum announced in September the open-sourcing of its quantum compiler Luoshu (LUOSHU), optimized for compilation of neutral-atom movement, dynamic connectivity and entanglement gate scheduling. At present, global quantum software roughly follows three routes. The first is led by hardware vendors, building ecosystems through proprietary SDKs, cloud platforms and compilers; the second pursues cross-hardware abstraction, allowing code to migrate between different QPUs and simulators; the third builds solvers in reverse from industry problems, where vertical companies in chemistry, finance and other fields compete.
Giants are equally active in this layer. Currently, IBM's Qiskit, Google's Cirq and NVIDIA's CUDA-Q are all building open quantum software ecosystems, connecting developers, algorithms and underlying hardware.
The three advances ultimately need to be combined into one system. At present, Taichi Quantum is advancing system-level integration of the physical platform, real-time control and quantum compilation software, striving to form an engineering system with automatic error correction capability by the end of the year.
The other side of pragmatism is that Taichi Quantum is actively exploring application scenarios. On one hand, growing AI demand has brought computing power anxiety; on the other hand, there is the application of quantum computing in vertical industries. According to the company, Taichi Quantum has cooperated with more than ten customers in industries such as nuclear industry, energy and chemicals, new energy and new materials, biomedicine, and finance around real-scenario problems, accelerating the industrialization of quantum computing.
Fang Zhenghao believes that the industrialization of quantum computing requires the joint promotion of technological breakthroughs and customer demand. On one hand, the company at this stage values actual revenue generated from scientific research and industrial cooperation; more importantly, it uses real scenarios to verify demand, accumulate industry experience, establish long-term customer relationships, and lay a commercial foundation in advance for the large-scale application of fault-tolerant quantum computing in the future.
But it cannot be denied that as a frontier industry, quantum computing has not yet converged on technology routes, and there are also disagreements on commercialization timelines. Regarding the realization of quantum computing, pessimists believe it is far off, while optimists believe it will happen within five to ten years. For example, in early 2025, Jensen Huang expected quantum technology to take 20 years to reach practical application, but less than half a year later changed his tune, saying quantum computing is ushering in an "inflection point" and will solve some "interesting global problems" in the next few years. Accordingly, NVIDIA has actively positioned itself in recent years, investing in leading quantum companies such as Quantinuum and PsiQuantum, and also releasing quantum computing open-source models and CUDA-Q, attempting to become the "operating system" supplier of the quantum computing era.
The phenomenon of capital cycles preceding industry cycles also exists. For example, two overseas quantum computing companies listed on U.S. stocks. In June this year, trapped-ion quantum computing company Quantinuum listed on Nasdaq with a market value of more than 15 billion dollars, becoming the largest IPO in the history of quantum computing. In August this year, French neutral-atom quantum computing company Pasqal also officially listed on Nasdaq through a SPAC.
Another undeniable point is that domestic quantum computing entrepreneurship is indeed the result of the combined effect of policy, technological progress and capital breakthroughs, carrying a strong flavor of a "contested strategic ground," as major countries continue to increase their bets in the quantum field with policy and capital.
According to CVSource data from Touzhong Jiachuan, the number of domestic quantum technology investments in 2025 tripled compared with 2024. According to investors, the "valuations of quantum computing companies are doubling every few months."
The key turning point cannot be separated from the 15th Five-Year Plan, and continued technological progress is also a significant boost. Quantum computing has higher technical thresholds and scarcer matching talent, and the number of teams with real technical capabilities is limited. Even so, conservatively estimated, more than 20 neutral-atom startups have been born in the past year, with more diverse backgrounds including academia, returnees and engineering professionals.
Taichi Quantum is one of the particularly noteworthy samples. This company is a firm optimist. According to Fang Zhenghao's expectations, large-scale quantum computing is expected to move toward commercial application in the next five years and empower artificial intelligence (Quantum for AI) to accelerate development.
All of this explains why Taichi Quantum chose a more concentrated and even slightly aggressive investment approach: targeting global giants, building a team of nearly 100 people, advancing full-stack self-development of core technologies, and prioritizing resources for universal fault-tolerant computing. Through exploration around scenarios such as chemical reactions, material simulation and energy scheduling, it stands at a sufficiently frontier position to understand in advance what tasks future quantum computing can actually accomplish.
Comments