A New Approach to 'Measurement,' the Bottleneck of Quantum Computing

SCSK Corporation and Quemix Inc. have developed "POD Readout," a new technology (patent pending) that resolves the "readout" bottleneck hindering the practical application of quantum computers. This technology is based on the concept of directly extracting only the information truly important to the solution, rather than reading all the vast calculation results generated by the quantum computer. This reduces the number of measurements by up to 1/1000, paving the way for practical simulations and analysis in industrial fields such as manufacturing, materials, and finance without compromising the high-speed nature of quantum computing.

1. Social Background and Challenges
In modern society, advanced simulations and analysis are essential for achieving carbon neutrality by 2050 and managing risks in complex financial markets. As the scale and precision of these tasks increase, the computational load grows exponentially. Since traditional computers are reaching their limits, expectations for quantum computers as next-generation accelerators are rising.

While quantum computing power has improved, "post-calculation readout (measurement)" has become a major challenge. Although quantum calculations themselves are fast, the cost of accurately extracting results as numerical values creates a significant bottleneck that can negate the speed of the quantum computer.

2. Overview of "POD Readout"
Since their capital and business alliance in 2024, SCSK and Quemix have been conducting joint research. They developed "POD Readout" to efficiently extract important information from quantum calculation results, focusing on fluid dynamics. Unlike conventional methods that read quantum states sequentially, this technology constructs a reconstruction filter (POD basis) from classical numerical fluid calculations. By embedding this into the quantum circuit, it extracts only the important features, allowing for efficient reconstruction of the quantum state.

3. Contribution to Practical Quantum Computing
This technology has succeeded in reducing the number of measurements by up to 1/1000. It is highly compatible with applications requiring the extraction of key indicators and values, with expected uses in manufacturing, materials, finance, and digital twins.

4. Future Outlook and International Conference
Moving forward, the companies will research generalization performance to enable high-precision readout with fewer measurements even for unknown bases. They plan to present these research results at "Q2B 2026 Tokyo," held on June 4-5, 2026.