RIKEN Launches ROQUO Supercomputer to Drive Full-Scale Quantum-HPC Integration in Japan

[Current]

RIKEN, Japan's flagship national research institute, launched the ROQUO supercomputer specifically designed to enable tight integration between quantum processors and high-performance classical computing resources. The system is intended to act as the classical control and interface layer for quantum hardware, handling the enormous classical computation required to run error correction algorithms, manage qubit calibration, and process quantum measurements in real time. RIKEN has been operating one of Japan's leading quantum computing programmes and ROQUO represents a significant infrastructure commitment to moving beyond individual quantum experiments toward hybrid quantum-classical workflows at research scale.

The quantum-HPC integration problem is often underappreciated in coverage of quantum computing. A quantum processor running at near absolute zero cannot directly communicate with room-temperature classical systems through conventional electronics without introducing thermal noise that destroys quantum states. The interface layer requires specialised cryogenic electronics, high-speed classical processing for real-time error correction, and software that can orchestrate mixed quantum-classical algorithms. ROQUO is designed to sit in that interface role.

Japan's quantum computing strategy has emphasised Rapidus, the government-backed foundry targeting 2nm semiconductor manufacturing by 2027, and quantum computing as a parallel track of domestic capability building. IBM mentioned Rapidus specifically when discussing which foundries might adopt the nanostack architecture that underlies its 0.7nm chip research. The ROQUO launch suggests Japanese quantum research infrastructure is advancing on both the hardware and systems integration sides simultaneously.

RIKEN Launches ROQUO Supercomputer to Drive Full-Scale Quantum-HPC System Integration - Quantum Computing Reportquantumcomputingreport.com

[ShawnMichaels99]

The classical control problem for quantum computers is not discussed nearly enough. A quantum processor with 1000 physical qubits needs more classical compute for real-time error correction than most organisations have in total

[Rocket67]

RIKEN building dedicated HPC for quantum integration rather than trying to run quantum experiments alongside conventional HPC workloads on shared infrastructure is the right engineering decision

[Blake_32]

IBM mentioning Rapidus as a potential foundry partner for nanostack and RIKEN building quantum control HPC in the same week suggests Japan's semiconductor and quantum strategies are converging in interesting ways

[Sigma]

Real-time error correction for superconducting qubits requires microsecond latency between the quantum processor and the classical control system. You cannot do that over a network. ROQUO being dedicated infrastructure adjacent to the quantum hardware matters

[ShawnMichaels]

Japan has been funding quantum research seriously but quietly. The Moonshot Programme, the RIKEN quantum centre, now ROQUO. They are building a comprehensive national capability rather than just buying IBM cloud time

[Taker92]

The hybrid quantum-classical workflow is where the first useful quantum applications will emerge. Problems that are too large for pure quantum but benefit from quantum subroutines handling specific components. ROQUO seems designed for exactly that research agenda

[Amber Drifter]

I am curious about the connectivity between ROQUO and the actual quantum hardware. The bandwidth and latency specs for the quantum-classical interface will determine what kinds of error correction protocols are practically feasible