Compiler design for hardware specific decomposition optimizations, tailored to diamond NV centers

TL;DR

The paper addresses the challenge of translating quantum circuits into hardware specific instructions for diamond NV center based quantum hardware. It introduces a hardware aware compiler built on Qiskit that leverages diamond specific operations such as direct carbon control and partial swaps, and integrates classical tasks like state tomography. Through simulations on a diamond NV center oriented simulator, the work demonstrates reduced noise and improved fidelity due to hardware aware decompositions, validated by teleportation based CNOT and GHZ tests. This approach enables higher fidelity circuit execution on NV center platforms and provides a foundation for extending hardware specific compilation to other quantum technologies.

Abstract

Advances in quantum algorithms as well as in control hardware designs are continuously being made. These quantum algorithms, expressed as quantum circuits, need to be translated to a set of instructions from a defined quantum instruction-set architecture (ISA), which are executed by the control hardware. These translations can be done by a compiler, targeting different qubit technologies. Specifically for diamond NV centers, no compiler exists to perform this translation. Therefore, in this paper we present a compiler designed for quantum computers utilizing diamond NV center specific instructions, such as direct carbon control and partial swaps, to reduce execution times and gate count. Additionally, our compiler adds on top of general compilers by allowing classical instructions to perform state tomography and measurement-based operations. The output of the compiler is tested in a diamond NV center specific simulator. Comparing a general compiler output with the diamond NV center specific output of our compiler while applying decoherence and depolarization noise showed reduced noise effects due to diamond specific decomposition. The compiler was also tested to perform state tomography and measurement-based operations, which showed to be functional. Our results show that we have successfully created a compiler with integrated classical and quantum instructions support, which can improve circuit execution fidelity by utilizing diamond specific optimizations.

Figures (15)

• Figure 1: A visual representation of the qubits within a diamond NV center. Presenting both the electron and carbon qubits in the system. In this system 2 qubit gates are possible from electron on carbon and entanglement generation is possible between electrons on different NV centers.
• Figure 2: A representation of compiler stages, indicating the functionalities of the compiler. The stages presented in green are added by us, while the stages in white are already present in Qiskit.
• Figure 3: Quantum circuit depicting the decomposition of a cx gate from a carbon onto the electron in the native gateset of diamond NV centers. The left hand side of the figure shows the instruction to be executed, while the right hand side shows what operations are needed to realise the instruction in diamond NV centers.
• Figure 4: Quantum circuit representing an single qubit rotation gate on the carbon when the electron state needs to be preserved (DDrf gate). The left hand side shows the operation that needs to be executed, while the right hand side shows a solution to perform the instruction on diamond NV centers while preserving the state of the electron.
• Figure 5: Quantum circuit representing a single qubit rotation gate on the carbon when the electron state does not need to be preserved (direct control). The left hand side shows the operation that needs to be executed, while the right hand side shows a solution to perform the instruction on diamond NV centers while destroying the state of the electron, but reducing the amount of instructions needed.