Forensics of Transpiled Quantum Circuits
AI Breakdown
Get a structured breakdown of this paper — what it's about, the core idea, and key takeaways for the field.
Abstract
There is a steady growth in research and demand for real quantum hardware in the Noisy Intermediate-Scale Quantum (NISQ) era of quantum computing. This prompts many third-party cloud providers to set up quantum hardware as a service that includes a wide range of qubit technologies and architectures to maximize performance at minimal cost. Different backends vary in terms of noise behavior, the basis gate set, coupling architecture and speed, among other factors. However, there is little to no visibility on where the execution of the circuit is actually taking place. The success of the user program is highly reliant on the backend that was used for execution. Besides, the third-party provider and/or tools (e.g., hardware mapper and allocator) may be untrustworthy and execute the quantum circuits on less efficient and more error-prone hardware to conserve resources and maximize profit. As such, gaining visibility on the backend from various aspects of the computing e.g., transpilation, execution and outcomes will be extremely valuable. Towards this goal, we introduce the problem of forensics in the domain of quantum computing where the objective is to trace the hardware and its properties where the quantum program has been transpiled and executed. Effective forensics can have many applications including establishing trust in the quantum cloud services. In this work, we focus on tracing the coupling map of the hardware (from the transpiled program) where the transpilation of the circuit took place. Next, we extracted the coupling map of the whole backend using as little as 3 transpiled circuits, both for user-based and transpiler-based (using auto-assignment feature) logical to physical mapping of the qubits. Finally, we addressed the problem of tracing the backends in a suite of backends using a pool of transpiled quantum circuits of varied sizes.