Logical Qubit Technology Opens a Superconducting Quantum Cloud Platform

Logical Qubit Technology, a Chinese quantum computing company, has launched a cloud platform that gives remote access to its superconducting quantum chips. Users submit their algorithms as code, and the platform runs them on the company’s hardware and returns the results, so researchers can use a quantum computer without buying or maintaining one. The company unveiled it at its Quantum Day event on July 9.

What Logical Qubit Technology’s cloud platform offers

Per a July 14 report, access runs across the company’s AGate chips, including 30-qubit and 100-qubit systems. Logical Qubit Technology reports single-qubit gate fidelity up to 99.97% and two-qubit fidelity up to 99.8%, with median parallel figures on the AGate-100 somewhat lower. These are the company’s own numbers, not independently benchmarked, so its claim that they meet or exceed leading international platforms should be read as a company assertion.

More concrete is the tooling for error-correction research. The platform lets users run distance-9 repetition code and distance-3 surface code, the kind of experiments needed to test how logical qubits suppress errors. It also offers pulse-level control, meaning advanced users can shape the actual waveforms sent to each qubit down to one-nanosecond timing, as opposed to working only with standard gates. The company frames this as running a “quantum computing lab” online.

The roadmap from cloud access to fault tolerance

Logical Qubit Technology announced several partners, the most notable being a strategic agreement with the pharmaceutical services firm Pharmaron to explore quantum applications in drug research through joint labs. Several software and algorithm companies are also on the platform.

The company laid out an aggressive roadmap: crossing the fault-tolerance threshold in 2027, producing results that beat classical computers in areas like drug discovery by 2028, and reaching tens of thousands of physical qubits by 2029. Those are targets, not results. It also reported a coherence time above 890 microseconds on a small test chip, with the next step being to move that onto its larger processors. Whether the roadmap holds is the open question.