Amazon Braket now supports parametric circuits to speed up variational algorithms
Amazon Braket, the quantum computing service from AWS, adds support for parametric circuits on superconducting, gate-based quantum processing units (QPUs), enabling up to 10X faster runtime performance for quantum-classical algorithms using Amazon Braket Hybrid Jobs.
Amazon Braket, the quantum computing service from AWS, adds support for parametric circuits on superconducting, gate-based quantum processing units (QPUs), enabling up to 10X faster runtime performance for quantum-classical algorithms using Amazon Braket Hybrid Jobs.
Parametric quantum circuits, which use gates with adjustable parameters, enable researchers to speed up variational algorithms such as quantum machine learning and quantum approximate optimization algorithms by reducing the overhead associated with the computationally intensive compilation step. With this launch, customers can submit parametric circuits, and Braket will compile the circuit once without the need to recompile for subsequent parameter updates to the same circuit. This is designed to remove compilation from the critical path, improve runtime performance by up to 10X for algorithms submitted as a Hybrid Job, and allow customers to take maximum advantage of the QPU for their research.
Parametric circuits are now supported at no additional cost on all QPUs from Rigetti Computing and Oxford Quantum Circuits in the US West (N. California) and Europe (London) AWS Regions respectively. To get started, check out our documentation and example notebooks here: