Hybrid quantum computing refers to processes and architectures that mix classical and quantum computing, enabling both kinds of systems to work together to solve a problem. Classical computers have always been used in quantum computing to define quantum gates, control configuration of the quantum computer, handle job submission, and to process results from the quantum computer. But with the latest generation of “integrated hybrid” quantum computing architectures – available now in Azure Quantum – quantum application developers can mix classical and quantum programming instructions in a single application, and can adjust quantum circuits while qubits remain coherent. 
 

As quantum technology evolves and advances, classical and quantum processes are becoming increasingly integrated. Microsoft has developed a 4-stage taxonomy to define increasingly sophisticated levels of hybrid quantum computing.

• Batch quantum computing
  Batch quantum architecture uses local clients to define circuits, and then submit the circuits sequentially as jobs to the quantum processing unit (QPU), which then returns the result to the client. Batching multiple circuits into one job eliminates the wait between job submissions to the QPU, enabling faster execution of the overall program. Examples of problems that can take advantage of batch quantum computing include Shor’s algorithm and simple quantum phase estimation.
  
  
• Interactive quantum computing   
  In this model, the client compute resource is moved to the cloud, resulting in lower latency between the client and QPU. This enables repeated execution of the quantum circuit with different parameters. Jobs can be grouped logically into one session and prioritized as a group on the quantum computer. Although sessions allow for shorter queue times and longer running problems, qubit states don’t persist between jobs. Examples of problems that can use this approach are variational quantum eigensolvers (VQE) and quantum approximation optimization algorithms (QAOA).
  
  
• Integrated quantum computing
  With integrated quantum computing, classical and quantum architectures are tightly coupled, allowing classical computations to be performed while physical qubits are coherent. Although it is still limited by qubit life and error correction, this architecture is a significant evolution in that it allows quantum programs to move away from just circuits. Quantum programs can now use common programming constructs such as loops to perform mid-circuit measurements, optimize and reuse qubits, and adapt circuits in real-time. Examples of scenarios that can take advantage of this model are adaptive phase estimation and machine learning.
  
  
• Distributed quantum computing
  Distributed quantum computing architectures will be unlocked when we have scaled, quantum computers with robust error correction, logical qubits, and long qubit lifetimes. In this architecture, classical computation will work alongside logical qubits. The long coherence time provided by logical qubits will enable complex and distributed computation across heterogenous cloud resources such as HPC, AI, and QPUs. By paring QPUs composed of large number of qubits with distributed and powerful cloud resources, we expect this architecture to enable solutions such as the evaluation of full catalytic reactions. Unlocking this architecture is expected to yield vast commercial benefit, enabling applications that can be used to solve some of the hardest problems facing humanity, such as carbon capture and discovery of new drugs.

For more information on hybrid quantum computing, check out our recent Q# blog announcing the release.