NVIDIA enters the quantum computing field
Since Jensen Huang changed his attitude towards quantum computing from "skeptical" to "supportive", the tech circle has been completely abuzz.
Recently, NVIDIA, under Jensen Huang's control, has officially entered the quantum computing field. Reports indicate that its venture capital arm has made its first investment in Quantinuum, a quantum computing company majority - owned by Honeywell International Inc. This investment values the company at $10 billion.
Previously, during his speech at GTC Paris, Jensen Huang announced that NVIDIA was launching an open - source quantum development platform, CUDA - Q, aiming to be the "super connector" between quantum computing and classical high - performance computing.
Regarding NVIDIA's series of moves, some believe that with NVIDIA's market value exceeding $4 trillion, this giant is not only the world's largest company but is also moving towards the throne in the new computing era by laying out the next - generation computing platform of "quantum computing + AI".
Jensen Huang Once "Poured Cold Water" on Quantum Computing
What attracted Jensen Huang to invest this time is Quantinuum, a quantum computing company majority - owned by Honeywell International Inc.
The company was established in 2021 through the merger of Honeywell's Quantum Solutions division and Cambridge Quantum Computing. It focuses on ion - trap quantum computing technology and has completed multiple rounds of financing. When it completed a $300 million financing in January 2024, its valuation was $5 billion. Compared with the beginning of 2024, it means that in just 18 months, Quantinuum's financing valuation has doubled.
Jensen Huang once had a "skeptical" attitude towards quantum computing. In January 2025, he publicly stated that a "very useful" quantum computer might take decades to emerge, which directly led to a sharp decline in the stock prices of quantum - computing - related companies.
In his view, there are two core problems in quantum computing. On the one hand, the technology is extremely complex and is still in the basic research stage. Key technologies such as quantum error correction and qubit stability have not been broken through, and large - scale commercial applications face numerous obstacles. On the other hand, in the short term, quantum computing is difficult to replace classical computers. Its application scenarios are limited, and its performance - to - cost - effectiveness ratio lacks competitiveness.
Therefore, Jensen Huang questioned whether quantum computing could be commercially implemented within 15 years and believed that 20 years was more in line with the industry's actual expectations.
However, the change in attitude occurred in March 2025. At the "Quantum Day" event of NVIDIA's GTC technology summit, Jensen Huang publicly apologized, admitting that his previous timeline estimate for quantum computing was incorrect and officially retracting his "pessimistic" remarks. His attitude shifted from "skeptical" to "supportive".
In action, NVIDIA also responded quickly and participated in the leading company in the industry. In Quantinuum's current $600 million financing round, NVIDIA, Quanta Computer, etc., joined as new investors, while existing shareholders such as JPMorgan Chase, Mitsui & Co., and Amgen increased their investments.
After the new round of financing was completed, Quantinuum said that the funds would be used to accelerate the R & D progress to launch the new - generation quantum computing system Helios later in 2025 and to promote breakthroughs in general fault - tolerant quantum computing technology. Meanwhile, Quantinuum plans to launch an IPO by the end of 2027.
Quantum Computing + AI: Is Jensen Huang Creating an AI "Nuclear Weapon"?
As is well - known, after achieving a dominant position in the AI computing power field, NVIDIA has been continuously expanding its business boundaries through investments. Currently, its influence has extended to multiple fields such as healthcare, artificial intelligence and cloud computing, and AI infrastructure.
Quantum computing is no exception. The CUDA - Q platform launched by NVIDIA in March this year is precisely Jensen Huang's core concept of "quantum computing + AI".
According to Jensen Huang's vision, in the past few decades, the world's most powerful computing power has been achieved by GPUs. However, in the future, every supercomputer will be equipped with a quantum processing unit (QPU) to work together with the GPU. Among them, the QPU focuses on quantum computing tasks, while the GPU undertakes pre - processing, control, error correction, and post - processing tasks.
The synergy between the two is like installing two engines in a car. Not only does it run faster, but it also changes the way of running. After changing the way of running, computing is no longer a "linear speed - up" but becomes a "quantum superposition".
Jensen Huang predicts that the growth of the number of qubits (the basic unit in quantum computing) in quantum computing will follow a pattern similar to Moore's Law: increasing by 10 times every 5 years and 100 times every 10 years. This exponential growth allows quantum computing power to quickly surpass traditional computing and directly enter the realm of complex problems that humans could not previously solve. For example, global financial system simulation, training of AI models with hundreds of billions of parameters, and interstellar - level physical modeling...
To promote the rapid implementation of quantum computing, Jensen Huang began to create the next - generation AI computing power "nuclear weapon" - the CUDA - Q platform. Past experience shows that NVIDIA's success does not only come from the hardware advantage of being the "computing power king" but also from its powerful software ecosystem.
In terms of computing power, AMD's MI300 released last year has exceeded NVIDIA's H100, and Intel's GAudi 3 is claimed to be 30% faster than NVIDIA's H200. However, these two giants cannot take away NVIDIA's market share. One of the core reasons is that NVIDIA has built an irreplaceable software ecosystem barrier through CUDA.
The CUDA - Q platform is the same. As an important extension of the CUDA platform, it follows the same idea: in the field of quantum computing, NVIDIA does not directly develop quantum processing units (QPUs). Instead, it relies on the CUDA - Q platform to integrate all the ecosystems around quantum computers, enabling developers to simultaneously mobilize the computing power of GPUs, CPUs, and QPUs in one program to achieve hybrid programming, thus providing a powerful tool for the integration of quantum computing and classical computing.
In addition to the CUDA - Q platform, NVIDIA also provides diversified services such as cloud services, AI infrastructure, electronic circuit simulation, and SDK to further improve the "quantum computing + AI" ecosystem.
Jensen Huang mentioned in a video interview that over the years, NVIDIA has built a close upstream - downstream cooperation ecosystem. Currently, many large enterprises have adopted NVIDIA's computing power platform, including giants such as Microsoft, Google, Meta, and Tesla.
After investing in Quantinuum, NVIDIA will deepen its quantum computing layout. First, the software tools (such as CUDA - Q) will synergize with Quantinuum's hardware to improve the quantum - classical hybrid computing ecosystem. Second, the two parties will jointly promote the establishment of NVIDIA's "Accelerated Quantum Research Center" in Boston, which is expected to be put into operation later in 2025 to promote the in - depth integration of quantum technology and AI technology.
Industry Giants Compete for the High - Ground in the Next Round of Technological Development
Actually, not only Jensen Huang, but in recent years, all major global technology - strong countries have attached great importance to quantum computing research and have issued quantum information technology strategies one after another, striving to seize the high - ground in the next round of technological development.
In addition to NVIDIA's investment in Quantinuum, another US quantum computing startup, PsiQuantum, also completed a $620 million Series D financing at the end of 2024. This round of financing was led by BlackRock, with the participation of the Australian federal government and the Queensland government. The US Department of Energy also signed a cooperation agreement with PsiQuantum.
In terms of the amount, this financing became the largest single - round financing in the quantum computing field in 2024, exceeding the annual total financing in the quantum computing field of most countries and regions.
PsiQuantum was founded in 2016 by founder Jeremy O'Brien and three scholars. In just 9 years, this company headquartered in Palo Alto, California, has grown into an enterprise with 350 employees, and its capital reserve may be comparable to the internal quantum computing projects of Google and IBM.
Almost at the same time when Quantinuum completed its financing, a quantum computing unicorn was born in Europe.
On September 3, 2025, IQM, a Finnish quantum computing startup, announced the completion of a $320 million financing. The new round of financing was led by US cybersecurity venture capital firm Ten Eleven Ventures, with Finnish venture capital firm Tesi following. Jan Goetz, CEO and co - founder of IQM, revealed that this capital injection allowed IQM to successfully enter the "unicorn" club, with the company's valuation exceeding $1 billion.
In addition, the stock price of IonQ, an important company in the US quantum computing field, has soared by about 500% in the past year, which fully reflects the market's recognition of its technology. The company aims to become the "NVIDIA in the quantum computing field".
In the second quarter of 2025, the New Jersey State Pension Fund, a large US pension fund, made an important investment adjustment: it cleared its holdings of Alibaba stocks and simultaneously established new positions in D - Wave Quantum and Rigetti Computing. Among them, D - Wave Quantum claimed to have achieved "quantum supremacy" in March 2025; Rigetti Computing is a company focused on scientific research and development. IonQ, whose stock price has risen by about 500%, was also favored by the fund and saw an increase in its holdings.
This series of actions means that the US capital market has set off an "investment boom" in the quantum computing field.
Statistics from "Information and Communication Technology and Policy" show that by sorting through venture capital databases such as CB Insights and Crunchbase, as of April 2025, the total number of US quantum - computing - related enterprises has exceeded 200.
Behind this is not only the gradually clear commercialization process of quantum computing but, more importantly, its biggest feature compared with traditional classical computers is its extremely fast computing speed.
Satya Nadella, CEO of Microsoft, said that a palm - sized quantum chip can solve problems that all the computers on Earth combined cannot solve at present. Take Google's Willow quantum chip as an example. It only takes 5 minutes to complete complex calculations that would take a traditional supercomputer 10²⁵ years to complete. Once humanity enters the quantum computing era, fields such as biomedicine and materials science will undergo a complete transformation.
In the field of AI, when quantum computing arrives, it will also surpass the current AI computing power system centered on GPUs. In particular, the training time of large AI models may be compressed from months to minutes, and the self - evolution of AI will achieve true exponential growth, triggering the "flywheel effect" and the "intelligence explosion".
Therefore, the computing power revolution triggered by quantum computing is very likely to push human technological development to a new height. For NVIDIA, which is already the king in the AI computing power field, early layout has become an inevitable choice. This layout will allow NVIDIA to continue to play the role of the "king of computing power" in the new round of computing power transformation.
This article is from the WeChat official account "Dongsi Shitiao Capital" (ID: DsstCapital), author: Chen Mei, published by 36Kr with authorization.