Rigetti Computing is opening their Quantum Cloud Service (QCS), which we reported on last September, for public beta testing. QCS will be using the hybrid quantum-classical architecture that Rigetti has been promoting and will use the new Aspen architecture QPUs (Quantum Processing Units) that Rigetti announced in August. Initial versions of Aspen used for this public beta of QCS are based upon a 16 qubit configuration. Subsequent versions of the Aspen family will extend to 128 qubits later in 2019 but it not yet clear if these higher density versions will be open to the public.
Users will receive $5,000 in QPU usage credits in the first month. This would be enough credits to provide over 5 hours of time on a full GPU chip. Two Aspen GPU’s will be available for this beta program and users will need to reserve time on one of the QPU’s using an online reservation system.
The QCS system is unique because each user gets a dedicated Quantum Machine Image (QMI) which is preconfigure with Rigetti’s Forest SDK. In addition, the QCS architecture has some unique features that are designed to specifically improved performance on hybrid algorithms. These include colocation of a classical CPU with the quantum processing unit to eliminate internet transmission delays, parametric compilation to allow running the same quantum program with different parameters at run-time, and an active qubit reset capability to allow for rapid execution of a programs eliminating any delays for a qubit to relax to the |0> state. All together, Rigetti states that these features can provide a performance increase of up to 30X compared to a web API model.
The Quantum Machine Images (QMI) will be preloaded with certain sample applications and Jupyter notebooks include programs for QCompress (quantum data compression), QClassify (facilitate classification to train variational quantum circuits), QuantumFreeze (a game), and Quantum Feature Detection (quantum machine learning functions) and others.
For more details, you can view Rigetti’s blog postings announcing the QCS public beta and providing additional details on their architecture. The blog titled Quantum Cloud Services Open in Public Beta can be found here, the blog titled Fast hybrid programming with Quantum Cloud Services is here, and the blog titled Meet Your Quantum Machine Image can be seen here.