YAQQ: Yet Another Quantum Quantizer - Design Space Exploration of Quantum Gate Sets using Novelty Search

Aritra Sarkar, Akash Kundu, Matthew Steinberg, Sibasish Mishra, Sebastiaan Fauquenot, Tamal Acharya, Jaroslaw Adam Miszczak, Sebastian FeldarXiv 2024 · DOI · arXiv

Introduction

In the race to build practical quantum computers, research often focuses on increasing qubit counts and improving coherence times. However, the efficiency of the "instruction set"—the fundamental native gates executed by a quantum processing unit (QPU)—is a critical yet often overlooked factor. This paper introduces YAQQ (Yet Another Quantum Quantizer), an open-source framework that utilizes evolutionary "Novelty Search" to discover optimized, hardware-specific gate sets that can significantly enhance performance in terms of circuit depth and fidelity.

The Challenge of Native Gate Sets

Quantum algorithms are typically expressed in high-level languages and must be decomposed into a native gate set supported by the target QPU. The selection of this set represents a vast design space. If a gate set is poorly matched to the target algorithms or hardware constraints, the resulting circuits become excessively deep. In the current Noisy Intermediate-Scale Quantum (NISQ) era, deeper circuits lead to increased decoherence and lower fidelity. While standard universal gate sets are theoretically sufficient, they are often practically inefficient for specific tasks such as quantum error correction (QEC) or specialized pulse-level control. YAQQ addresses this by treating gate set selection as a Design Space Exploration (DSE) problem.

Methodology: Novelty Search for Gate Discovery

The core innovation of YAQQ is the application of Novelty Search (NS) to quantum gate discovery. Unlike traditional optimization methods that pursue a single "fitness" goal—which can lead to stagnation in local optima—NS rewards solutions that are dissimilar to previously explored ones. This approach allows the algorithm to navigate the continuous, high-dimensional parameter space of quantum gates more effectively.

The YAQQ framework operates through three integrated components:

  1. Compiler: Decomposes target unitaries into a candidate gate set using techniques like Solovay-Kitaev (SKD), Cartan (KAK), or Quantum Shannon Decomposition (QSD).
  2. Comparator: Evaluates candidate gate sets against benchmark datasets.
  3. Discovery: Uses NS to iteratively refine gate sets based on a multi-modal cost function balancing process fidelity, circuit depth, and novelty.

Benchmarking with Diverse Datasets

To evaluate the performance of discovered gate sets, YAQQ utilizes several target distributions on the Bloch sphere. These include Haar random unitaries and states for general-purpose benchmarking, as well as structured datasets like equispaced angles and points for testing state preparation efficiency.

Visualizing target distributions on the Bloch sphere: (a) Haar random states, (b) Haar random unitaries, (c) Equispaced angles, and (d) Equispaced points.

Additionally, the framework incorporates QEC-specific states and the MQT Bench suite—a collection of 45 real-world quantum algorithms—to validate gate sets in practical scenarios.

Key Results

YAQQ has demonstrated superior performance compared to standard universal sets:

  • Enhanced Fidelity and Depth: The framework discovered 1-qubit gate sets that achieved higher average process fidelity and significantly lower circuit depth than the standard {H, T} set when tested on Haar random unitaries.
  • QEC Optimization: YAQQ successfully compared transversal gates for Reed-Muller and Steane codes, finding that Reed-Muller is more amenable to Solovay-Kitaev decomposition for magic states, while Steane performs better for cardinal states.
  • Scalability: Gate sets optimized on 1-qubit data were successfully scaled to decompose complex multi-qubit algorithms from the MQT Bench, maintaining their performance advantages.

Conclusion and Future Work

YAQQ highlights the importance of "reachability"—the efficiency with which a gate set can reach a target state within a hardware-imposed depth budget. By providing a tool for hardware-agnostic discovery, YAQQ offers a roadmap for designing QPUs with native instructions optimized for specific software stacks. Future work will focus on accelerating the DSE process using GPUs and extending the framework to account for noisy quantum compilation by representing operations as superoperators.

Summary updated

Abstract

In the standard circuit model of quantum computation, the number and quality of the quantum gates composing the circuit influence the runtime and fidelity of the computation. The fidelity of the decomposition of quantum algorithms, represented as unitary matrices, to bounded depth quantum circuits depends strongly on the set of gates available for the decomposition routine. To investigate this dependence, we explore the design space of discrete quantum gate sets and present a software tool for comparative analysis of quantum processing units and control protocols based on their native gates. The evaluation is conditioned on a set of unitary transformations representing target use cases on the quantum processors. The cost function considers three key factors: (i) the statistical distribution of the decomposed circuits' depth, (ii) the statistical distribution of process fidelities for the approximate decomposition, and (iii) the relative novelty of a gate set compared to other gate sets in terms of the aforementioned properties. The developed software, YAQQ (Yet Another Quantum Quantizer), enables the discovery of an optimized set of quantum gates through this tunable joint cost function. To identify these gate sets, we use the novelty search algorithm, circuit decomposition techniques, and stochastic optimization to implement YAQQ within the Qiskit quantum simulator environment. YAQQ exploits reachability tradeoffs conceptually derived from quantum algorithmic information theory. Our results demonstrate the pragmatic application of identifying gate sets that are advantageous to popularly used quantum gate sets in representing quantum algorithms. Consequently, we demonstrate pragmatic use cases of YAQQ in comparing transversal logical gate sets in quantum error correction codes, designing optimal quantum instruction sets, and compiling to specific quantum processors.
BibTeX
@misc{sarkar2024yaqq,
  author = {Aritra Sarkar and Akash Kundu and Matthew Steinberg and Sibasish Mishra and Sebastiaan Fauquenot and Tamal Acharya and Jaroslaw Adam Miszczak and Sebastian Feld},
  title = {{YAQQ: Yet Another Quantum Quantizer - Design Space Exploration of Quantum Gate Sets using Novelty Search}},
  year = {2024},
  eprint = {2406.17610},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2406.17610},
}
Figure