Put the wool and knitting needles down and get to Qiskit. Quantum Knitting is a way to make quantum circuits more efficient. Technically Circuit Knitting is decomposing a quantum circuit into smaller circuits and “knitting” those smaller results to build the outcome of the original larger circuit. Circuit Knitting has now come to the popular Qiskit Quantum Framework.
So what is Circuit Knitting?
Circuit Knitting decomposes a quantum circuit into smaller circuits, executes those smaller circuits on a quantum processor(s), and then knits their results into a reconstruction of the original circuit’s outcome. Circuit knitting includes entanglement forging, circuit cutting, and classical embedding. The Circuit Knitting Toolbox (CKT) is a collection of such tools now available to Qiskit.
Each tool in the CKT partitions a user’s problem into quantum and classical components to enable efficient use of resources constrained by scaling limits, i.e. the size of quantum processors and classical compute capability. It can assign the execution of “quantum code” to QPUs or QPU simulators and “classical code” to various heterogeneous classical resources such as CPUs, GPUs, and TPUs made available via hybrid cloud, on-prem, data centres, etc. The toolbox enables users to run parallelized and hybrid (quantum + classical) workloads without worrying about allocating and managing the underlying infrastructure, treating it as a black box.
To read more on the project go here.
