NTT tests a simpler software route to an optical quantum computer
The toolkit builds quantum circuits internally and connects to MoQuren hardware.
Reading tools for text size, copying, printing and your reading list

NTT has tested machine-learning software on an optical quantum computer without requiring developers to write quantum circuits. The November 16 announcement describes software that builds those circuits internally.[1]
Developers describe machine-learning tasks; the toolkit connects through OptQC’s software to MoQuren and returns results. NTT tested classification: assigning inputs to categories. Quantum calculations contributed to training the model and making predictions.[1]
The hardware predates this integration. OptQC announced MoQuren’s operation at Japan’s AIST research institute on July 21. Its launch announcement also described work on developer tools, including simulation capabilities.[2]
The demonstration establishes a software-to-hardware connection, not better machine learning. NTT supplies no dataset, accuracy score, conventional-computing comparison or public download date.[1]
For developers, the attraction is less circuit-writing before testing an idea. Whether the resulting application is useful still depends on its answers and the cost of obtaining them.