K-Semiconductor Legend Continues with Quantum... The Future Drawn by KAIST's 'Quantum Fab' [Paek Jongmin's Shockwave]

Supporting Technological Independence with the Establishment of Quantum Fabs
Semiconductors: The Foundation Behind Next-Generation Quantum Computer Technology 'Photonics'
Master Plan for Quantum Science, Technology, and Industry Development Sets Ambitious Goals for Core Technologies and Fostering Companies

On January 27, upon entering the laboratory of Professor Youngwook Sohn from the Department of Physics at the Korea Advanced Institute of Science and Technology (KAIST), where research on the core of quantum computers, the QPU, is underway, a phrase reflecting the earnest hopes of researchers for experimental success immediately caught the eye. This was an example of how challenging this research is.

Professor Youngwook Sohn from the Department of Physics at KAIST is unveiling the QPU, the core of optical quantum computers. Photo by Paek Jongmin, Tech Specialist

On one of the laboratory’s measurement devices, the waveforms of photon signals and electrical signals were being displayed in real time. Professor Sohn and his team held up a QPU board that combines a photonic chip and a semiconductor chip, explaining the experimental process in detail.

Professor Sohn explained on site that, “The photon-based approach is one of the most advantageous ways to scale up quantum computers.” While there are various qubit architectures such as superconductors, ion traps, and neutral atoms, he argued that although optics is a latecomer, it is a strong contender. The reasons are scalability and its relevance to Korea.

Photon-based QPU developed by Professor Youngwook Son's team. Photo by Paek Jongmin, Tech Specialist

According to Professor Sohn, quantum computers using superconducting or ion trap methods have developed by creating one qubit, then gradually increasing the number by bringing them closer together. As the number of qubits increases, physical constraints and error rates inevitably rise as well.

Professor Sohn explained, “The photon-based approach is fundamentally different.” He said, “Rather than stacking qubits one by one, the photon-based method builds the system around a device that continuously generates qubits.” He compared this to a car: the core device that generates photonic qubits serves as the engine, and the system’s capacity for handling qubits depends on the performance of this engine.

In fact, it is known that the world’s most advanced photonic devices can generate hundreds of millions, or in some cases, billions of photonic qubits per second from a single device. Professor Sohn noted, “We do not call this a quantum computer by itself,” explaining that “it only becomes meaningful when the rest of the system that processes quantum information is combined.” However, he emphasized that once such a structure is established, the number of qubits can be scaled up at a pace entirely different from conventional approaches.

Professor Sohn also mentioned his experience at PsiQuantum, a company where he previously worked. PsiQuantum is focused on photon-based quantum computing and is already constructing a computing facility the size of a soccer field. Their goal is to build a system with hundreds of thousands, or even up to a million qubits. Professor Sohn explained that while such a target may sound exaggerated by conventional quantum computer standards, it is a technically plausible scenario for photon-based architectures.

Professor Sohn unveiled a QPU that integrates a photonic chip and a semiconductor chip into a single system to the press. He explained, “The photonic chip handles quantum information, while the semiconductor handles digital control and signal processing,” adding, “This semiconductor is not just a supporting technology; its role in the overall system is by no means insignificant.”

This structure naturally aligns with Korea’s strong semiconductor industry. The recently spotlighted Co-Packaged Optics (CPO) technology in the semiconductor field shares many technical similarities with photon-based quantum computing. CPO is a next-generation chip interconnection method that packages a semiconductor chip and silicon photonics (technology that processes optical signals directly on the chip) together, enabling data transmission via light instead of electricity.

Professor Sohn explained, “In the AI semiconductor field, technologies that combine CPUs and GPUs with optical communication are already being used industrially,” adding, “These manufacturing techniques could be directly leveraged in photon-based quantum computing as well.”

He continued, “While overseas, design and manufacturing are typically separated, Korea possesses both design capabilities and manufacturing infrastructure,” emphasizing, “In this respect, photon-based quantum computing is a platform where our country can be competitive.”

The Korea Institute of Science and Technology (KIST) is also conducting active research in the field of quantum optics. In September of last year, KIST introduced photon-based simulators and integrated optical quantum device technologies, and demonstrated remote connectivity. A photon-based QPU was also released.

Photonic QPU developed by the KIST research team. Photo by Paek Jongmin, Tech Specialist

The key infrastructure for linking such research to industrial applications is the quantum fab. Younghyun Cho, Director of the Quantum Fab Lab at KAIST, explained, “Quantum technology is now moving beyond the problem of building individual devices in laboratories to challenges of manufacturing and systems engineering.”

Traditional photon-based quantum experiments typically involved placing numerous optical components on large optical tables. However, this method is vulnerable to environmental changes such as temperature and vibration, and it is difficult to consistently reproduce the same performance. Director Cho emphasized, “By miniaturizing onto chips and stabilizing through manufacturing processes, error control and large-scale scalability become possible.”

Younghyun Cho, Director of the Quantum Fab Research Center at KAIST, is explaining the plan for the National Quantum Fab Research Center. Photo by Paek Jongmin, Tech Specialist

Currently at KAIST, construction is underway for the nation’s largest ‘open quantum fab’ research building, with completion targeted for 2027. By 2031, approximately 45.1 billion won will be invested to establish more than 37 types of core equipment, making it a hub for the industrialization of quantum technology.

The quantum fab being promoted by KAIST is not a facility that fabricates devices on behalf of specific research teams. Instead, it aims to be an open infrastructure where university researchers, government institutes, and companies can directly carry out processes and accumulate process recipes and know-how. Since photonic chip processing is similar to existing semiconductor processing, there is significant potential for linkage with Korea’s semiconductor industry.

KIST is also reportedly in the final stages of establishing its own quantum fab, securing a system capable of in-house design, component manufacturing, and packaging of photonic QPUs.

The government has also recognized the importance of quantum fabs and has taken proactive steps to support them. Shim Juseop, Director at the Ministry of Science and ICT, identified the quantum fab as a core infrastructure in the comprehensive quantum plan announced on January 29.

The first Master Plan for Quantum Science, Technology, and Industry Development, which serves as the master plan for the quantum sector, includes goals such as making Korea the world’s number one quantum chip manufacturer and fostering 2,000 quantum companies by 2035.