CRAFT - Characterizing Realistic Architectures for Fault-Tolerance
Abstract
In order for a general-purpose quantum computer to have provable speedups over classical hardware, it is necessary to show fault-tolerance with error correcting codes. In a fault-tolerant architecture, the ability to implement increasingly longer and more complicated algorithms, with arbitrarily low logical error rates, becomes a function of the physical error rate, which determines the necessary code distance and, through it, the number of physical qubits required. However, a rigorous and complete fault tolerant implementation of quantum error correction has not been demonstrated. Therefore, the objective is to characterize, as much as possible, all the components of a fault-tolerant architecture and to evaluate the relative errors in the different subsystems. It is unclear how to relate these subsystem metrics to that of fault-tolerance on the whole system, though certainly having properly functioning subsystems is a necessary condition to fault-tolerance. Towards that goal there is a rich set of tools from the field of quantum characterization, validation and verification (QCVV); however, the majority of these protocols are biased toward gate performance and so must be augmented for the task of QCVV for fault-tolerant architectures. The CRAFT project -- "Characterizing Realistic Architectures for Fault-Tolerance" -- aims to develop hardware agnostic characterization methods for benchmarking the subsystem components of a fault-tolerant quantum computing architecture. This project will be a theoretical and experimental investigation into these new characterization techniques, extended from traditional QCVV protocols such as randomized benchmarking. The final goals will be a toolkit of methods which can be incorporated into open source software packages such as Qiskit.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Nov 21, 2020
- Source ID
- W911NF2110002
Entities
People
- David Mckay
Organizations
- Army Contracting Command
- International Business Machines Corporation (Armonk, NY)
- National Security Agency