Researchers from Amazon Web Services (AWS), Nvidia, Lawrence Berkeley National Laboratory, and NASA have proposed a model to evaluate the interaction between quantum processors and classical supercomputers.
The model helps determine when quantum equipment should be placed near high-performance computing infrastructure and when remote access via the cloud is sufficient.
Requirements Depend on Task Type
The researchers divided quantum-classical interaction into two levels:
- The first involves real-time equipment management, including physical qubit calibration and quantum error correction data processing.
- The second covers hybrid algorithms where a classical computer and a quantum processor alternately perform parts of the computation.
In error correction, the classical system analyzes control measurements and determines which errors need fixing. Processing must occur within intervals ranging from fractions of a microsecond to several microseconds. Such tasks require direct, low-latency connections, making remote access via a standard network unsuitable.
In the second case, latency affects the duration of the task but does not always determine the feasibility of executing the algorithm. Therefore, some tasks can be performed with cloud access to the quantum device.
Remote Access Not Suitable for All Algorithms
The authors divided the execution time of a hybrid algorithm into classical computations, quantum processor operations, and data exchange.
If most of the cycle is spent on information processing, reducing network latency does not significantly speed up the computation. If the algorithm requires frequent exchanges, a fast connection becomes critical.
The first example was sample-based quantum diagonalization (SQD), used in quantum chemistry. The authors used data from an experiment with IBM Heron’s 77-qubit processor. In this scenario, classical processing of results took most of the time. Communication overhead was about 0.0001, so remote access to the quantum processor was sufficient.
A different result was shown by the quantum-enhanced Markov chain Monte Carlo method (QE-MCMC), tested on a 10-qubit IBM device. The algorithm requires frequent sequential exchanges. With remote access, communication overhead exceeded computation time by approximately 1000 times.
However, a full-fledged supercomputer was not required. A small classical controller with low-latency connectivity was sufficient for execution.
Error Correction Will Increase Requirements
As the transition to logical qubits and fault-tolerant quantum computers progresses, the role of classical infrastructure will grow. The system must continuously process measurements, detect errors, and send corrective commands. The speed of the classical part will begin to dictate the pace of the quantum computer.
The authors anticipate that large fault-tolerant setups will require closer integration with high-performance computing systems. However, there is no universal architecture. Requirements depend on the algorithm, equipment characteristics, number of exchanges, and data volume.
The researchers suggest using the model for regular reassessment of infrastructure as quantum processors and computational methods evolve.
In July, Nvidia released the code for an AI module for pre-processing quantum error signals. In simulations, pairing with a classical decoder reduced logical error rates and accelerated processing.
