Creating the Next Industry with Quantum Computers - Fujitsu Global

Event Report | 2026-03-10 10 minute read Fujitsu, which is currently developing a world-leading 256-qubit superconducting quantum computer, has laid out an ambitious roadmap for the Quantum Age that is expected to arrive in the 2030s. However, the real breakthrough will not be in the number of qubits, but rather in the quantum computer’s underlying design philosophy.Access here to watch digest video: https://www.youtube.com/watch?v=iRQkPg9FaEI Computing Aiming at “Ultimate Perfection” The WIRED Futures Conference 2025 brought together innovators from Japan and around the world who are leading the effort to develop and implement technologies for the approaching Quantum Age of the 2030s, when AI and quantum computers are expected to converge. The opening day’s session featured two major players in the field: Keisuke Fujii, a leading Japanese quantum computer researcher and a professor at The University of Osaka, and Shintaro Sato, the Head of Quantum Laboratory at Fujitsu Research. These two discussed the outlook for quantum computer implementation in a talk entitled “Creating the Next Industry: Japan at the Forefront of Implementation.” Fujitsu began to conduct full-scale quantum research around 2020, making it a relative latecomer in this cutting-edge field. This year, however, the company has taken the industry by storm by unveiling a 256-qubit superconducting quantum computer that has achieved the world’s highest level of performance, and it is currently working toward the goal of developing a 1,024-qubit superconducting quantum computer by 2026.Looking back on the early days when a powerful strategy was essential, Sato shared his insights: “The person we were looking for as a partner had to be someone at the forefront of the industry and who could think about quantum computers.” That was why the company chose Keisuke Fujii. Fellow, SVP, and the Head of Quantum Laboratory, Fujitsu Research, Fujitsu Limited. He started working at Fujitsu in the same year and has remained with the company ever since. He has received several awards for his research, including the SAP (the Japan Society of Applied Physics) Fellow Award in 2018. Since 2021, he has also served concurrently as the Deputy Director of the RIKEN RQC-Fujitsu Collaboration Center.PHOTOGRAPHS BY COMURAMAITEXT BY AKIHICO MORI In 2020, the early frontrunners overseas were competing intensely to develop a next-generation NISQ (Noisy Intermediate-Scale Quantum) device, which was the initial goal at the time. Fujitsu, by contrast, was looking beyond the limited performance improvements that could be achieved with existing NISQ technology. Instead, the company was focusing on the next stage of research and development. Fujitsu has been a leader in developing computers for many years. Among its achievements are the supercomputer Fugaku. The company is currently pursuing technology that has the potential to overcome the consequences of the end of Moore's Law. Specifically, it is working towards the realization of a fault-tolerant quantum computer (FTQC), a goal regarded as the “ultimate perfection” in quantum computing.

Quantum Error Correction (QEC) is essential to achieving this goal, and so Fujitsu has been moving quickly and steadily to refine this technology, approaching the task not as a sprint but as a marathon.When Fujitsu started working on quantum computers, Google's “quantum supremacy” was generating worldwide attention, and the competition to find ways of using existing quantum machines was intensifying. Fujii, who reviewed a paper that demonstrated Google’s quantum supremacy, was right at the center of things. “I had just started up a company called QunaSys, which focuses on quantum algorithm research, and we were collaborating with a number of other companies on NISQ applications. I explained to Fujitsu that it was hard to partner with other companies on NISQ, and they responded by suggesting the idea that we undertake joint research on error correction technology. Error correction is an important technology for realizing the ultimate quantum computer, and I was surprised that no other companies at the time had such a long-term vision.” This was how the partnership between Fujitsu and Fujii to work on quantum error correction over the long term was born, and they began their collaboration towards the future goal of ultimately creating a system with one million qubits.With hindsight, it was clear that Fujitsu's long-distance strategy made a lot of sense. In 2024, Google released the Willow chip, which made quantum error correction using logical qubits a tangible reality. This quickly shifted the global focus onto FTQC. Today, everyone agrees on one thing: there can be no practical future for quantum computing without quantum error correction.In August 2025, Sato announced an ambitious roadmap. The goal is to achieve practical quantum computing by 2030, by running hundreds of logical qubits (groups of qubits capable of high-precision computation using quantum error correction technology) on a quantum computer equipped with more than 10,000 physical qubits.” A key feature is that it clearly delineates the stage of “implementing logical bits to achieve practical computational accuracy” with an eye toward FTQC, rather than simply increasing the number of qubits. Research and development in large-scale cryogenic refrigeration systems, which is the biggest remaining challenge, has also begun in Japan through a NEDO project. Large-scale cooling equipment is being developed, and infrastructure development is progressing with a view toward the 2030 target. Fujitsu's quantum computer development roadmap presented at the event. The STAR Architecture Quantum computing needs quantum error correction in order to have a future. Where can we expect to make a breakthrough that will open the door to that future? Fujii points to “foundational software.” To make quantum computing truly functional, it is essential to focus on more than just the performance of the hardware. We also need to design the operating hardware, foundational software, and applications so they work in an organically integrated manner.“Within the overall structure, the middle layer is the foundational software. It interacts with both the hardware and the applications, which gives it an enormous impact. The STAR architecture (Space-Time efficient Analog Rotation quantum computing architecture) we are currently proposing is a technology that will greatly enhance the efficiency of the key processes in FTQC, and it represents a new trend in this field,” he explained.The STAR architecture is an architectural design concept for quantum computers that will dramatically improve how efficiently they can perform calculations. It is essential for realizing practical quantum computing on quantum computers using as many as 10,000 physical qubits, which are expected to appear by around 2030. Professor at The University of Osaka and Deputy Director of the Center for Quantum Information and Quantum Biology (QIQB). Quantum computer researcher. His publications include kyōi no ryōshi konpyūta (The Amazing Quantum Computer) (Iwanami Shoten) and kyōyō to shite no ryōshi konpyūta' (Quantum Computers as a Liberal Arts Subject) (Diamond). While qubits act as a source of quantum acceleration, a major reason why an enormous number of qubits are considered necessary to realize FTQC is because of the existence of special quantum states known as magic states. Magic states are easily disturbed by noise and they are subject to numerous errors during their generation process. To overcome this hurdle and improve the purity of magic states, a process called magic state distillation is essential. This distillation process is extremely resource-intensive, consuming a significant portion of a system’s overall computational resources. Improving the efficiency of this step is a bottleneck for the entire FTQC process, conventionally requiring an enormous amount of computational power, on the scale of one million qubits. “The design philosophy that optimizes this high-cost magic state distillation is precisely the STAR architecture,” Fujii explained.The STAR architecture proposed by Fujii and Fujitsu makes it possible to implement phase rotation gates—which are quantum logic gates that perform fundamental quantum computation operations and are essential for complex computations—using fewer resources. The STAR architecture has already shown the potential for performing practical error-corrected computing at the 60,000-qubit scale in the materials science field, for example, while reducing computational costs by at least an order of magnitude. As a result, practical algorithms, such as those for estimating the energy of matter, which previously required 1 million qubits, will be solvable in about 10 hours.Many people are only interested in big numbers such as so many tens of thousands of qubits. “However, the reality is that the key to success lies in achieving practical quantum computing,” Sato states confidently. “If we can develop a technology that accelerates calculations exponentially compared to conventional computers while using fewer qubits, then everyone will benefit.” Once Technology Takes Shape, It Quickly Comes into Bloom The STAR architecture’s true value lies in its versatility. It embodies a hardware-independent design concept that can be applied to any type of quantum computer hardware. “Until now, there has been no intermediate step between NISQ machines with around 100 qubits and ideal machines with a million qubits. Through the STAR architecture, we have created a practical middle ground, namely Early-FTQC with around 10,000 qubits," Fujii explained. Thanks to the STAR architecture, quantum computers are rapidly advancing and transcending the limitations imposed by different hardware approaches.Quantum computers have a wide range of potential applications, including in materials design, financial optimization, drug discovery, and climate modeling. While most current applications remain at the experimental NISQ stage, once they progress to the practical stage with the addition of quantum error correction, their computational accuracy and search efficiency will improve dramatically. “Researchers can’t yet predict the full value of quantum computers. The situation is strikingly similar to that of AI in its early days,” Fujii stated. “Research using neural networks already existed before ChatGPT, but at that time its applications had not yet become clear. The moment a general-purpose chat interface appeared, it quickly became integrated into our daily lives.”AI gained societal acceptance not because of the technology itself, but because it became available in a form that anyone could use. “The ‘ChatGPT phase’ for quantum computers is still some way off. But it is clear that by 2030, the advent of high-performance machines will give rise to applications that we cannot yet anticipate. For AI, the year 2000 is happening right now,” Fujii observes, keeping a watchful eye on that future.

Related Information Fujitsu Quantum Fujitsu's quantum computer development strategy, main technological approaches, potential applications in industrial sectors, and support for application development with companies will be systematically introduced. Unraveling the Latest Trends in Error Correction and Error Mitigation in Quantum Computers In recent years, the advancement of quantum computers has been accelerating. Fujitsu, in collaboration with RIKEN, has started providing the first 64-qubit superconducting quantum computer by a Japanese company in 2023.

Fujitsu Research Portal The Fujitsu Research Portal introduces the details of each technology in our research and development.