Building Domain-Specific Compilers Using Parameterized Circuit Instantiation
ORAL
Abstract
Parameterized circuit instantiation is a common technique encountered in the generation of circuits for a large class of
hybrid quantum-classical algorithms, as well as in quantum compilation infrastructures such as IBM Qiskit and Google Cirq.
Newer synthesis-based compilers such as BQSKit make heavy use of ansatz instantiation using numerical optimization.
Although ubiquitously deployed, instantiation has not been thoroughly studied in the context of circuit generation and optimization pipelines.
In this work, we will survey the methods and uses for circuit template instantiation, and show how to build domain-specific circuit processing workflows.
Examples include the QPredict synthesis algorithm. Quickly built on top of existing instantiation methods, it is able to synthesize hamiltonian simulation
circuits far faster and with better quality than existing general-purpose synthesis methods such as QSearch and QFAST.
hybrid quantum-classical algorithms, as well as in quantum compilation infrastructures such as IBM Qiskit and Google Cirq.
Newer synthesis-based compilers such as BQSKit make heavy use of ansatz instantiation using numerical optimization.
Although ubiquitously deployed, instantiation has not been thoroughly studied in the context of circuit generation and optimization pipelines.
In this work, we will survey the methods and uses for circuit template instantiation, and show how to build domain-specific circuit processing workflows.
Examples include the QPredict synthesis algorithm. Quickly built on top of existing instantiation methods, it is able to synthesize hamiltonian simulation
circuits far faster and with better quality than existing general-purpose synthesis methods such as QSearch and QFAST.
–
Presenters
-
Ed Younis
Lawrence Berkeley National Laboratory, Lawrence Berkeley National Lab
Authors
-
Ed Younis
Lawrence Berkeley National Laboratory, Lawrence Berkeley National Lab
-
Costin C Iancu
Lawrence Berkeley National Laboratory