Oxford Quantum Circuits (OQC) has published its development roadmap for the next nine years with an end goal of producing a quantum processor, codenamed Atlas with 50,000 logical qubits by 2034, the largest number that anyone else has disclosed to date. Leading up to the Atlas, the company will be developing a series of three additional machines, codenamed Genesis, Titan, and Athena with increasing qubit count and error rate. The company expects that the machine slated for 2028, Titan, will be able to provide users with a commercial advantage over classical alternatives on some applications.
A key technology element that they will be using for this is called the multi-mode dual-rail Dimon qubit. This method allows them to encode a logical qubit in the single-photon excitation subspace of the dimon, using the method of dual-rail encoding and allows them to reduce the Physical-to-Logical ratio of qubits by a factor of about 10X. An additional technology the company plans to lever is their coaxmon technology which implements the transmon qubit with a coaxial geometry with the qubit on one side of the substrate and the readout resonator on the other side of the substrate aligned to the qubit. To accomplish this, they will be integrating a through-sapphire substrate machining process. Recently, the company made a presentation at the APS Global Physics Summit describing how they were using this technology to create a two-qubit CZ gate that operated in 25 nanoseconds with a fidelity of 99.8%.
What is notable about this roadmap is the small system sizes they project. Titan with 200 logical qubits will fit into a single 100 mm wafer, Athena with 5,000 logical qubits is expected to fit into a single 200 mm wafer, and Atlas with 50,000 qubits is expected to fit into two 300 mm wafers that are capacitively coupled. The company is also projecting very low logical error rates with the Atlas ultimately achieving a 10-12 error rate using an advanced multi-mode qubit along with a supplemental outer code redundancy.
Of course, it is one thing to create a roadmap on paper and then another to successfully execute against it on schedule. Superconducting quantum computers have a number of challenges that need to be addressed including areas such as minimizing error rates, maintaining adequate cooling for a large number of qubits in a dilution fridge, routing the wires needed to control large numbers of qubits, implementing error detection and correction in the shortest possible cycle time, and much more. In addition, we are always concerned about what we call the “Unknown unknowns” which are problems that were not anticipated. So, we wish the team at OQC the best of success with their efforts and if they succeed, they will have a very competitive quantum system that can great solutions for difficult problems in financial services, national security, and many other areas.
For more about OQC’s roadmap, view the press release and a roadmap presentation on their website here.
June 5, 2025
Leave A Comment