#quantumcomputingchallenge

Fault-tolerant Clifford Gates

Researchers Reveal Fault-Tolerant Clifford Gates Methods in Quantum Computing Revolution

A team of National Physical Laboratory and University College London researchers described new algorithms for implementing fault-tolerant logical Clifford gates on stabiliser quantum error-correcting codes, advancing the search for reliable quantum computers. This idea addresses qubit, gate, and measurement noise and errors, which prevent large-scale quantum computers from being built.

Challenge of Quantum Computing

Quantum computing's potential lies in its capacity to execute sophisticated calculations that traditional computing cannot. However, imprecise processes and outside noise quickly destroy quantum information, making quantum systems unstable. Quantum error correction (QEC), which maintains quantum information over many physical qubits, is a fantastic way to fix these issues and keep data.

Only protecting encoded data is insufficient. Fault-Tolerant Clifford Gates are needed for efficient logical operator operations on encoded data in quantum algorithms. Fault tolerance allows logical errors to be repaired and prevented from spreading. We need this for quantum computations to work. Clifford gates are essential to quantum computing and many quantum algorithms. Hadamard, S, CNOT, and CZ gates perform these operations. It's difficult to implement fault-tolerant logical Clifford gates without low-depth circuits.

Leveraging Symmetries: A New Method

Hasan Sayginel, Stergios Koutsioumpas, Mark Webster, Abhishek Rajput, and Dan E. Browne used stabiliser code symmetries to design a rigorous method. Their primary idea is to convert a stabiliser code to a conventional binary linear code and find the automorphism group, the set of bit permutations that preserve the code. Real circuits with fault-tolerant logical operations are created from these permutations.

“Our algorithms provide a rigorous formulation for finding automorphisms of stabiliser codes,” says corresponding author Hasan Sayginel. One of their main tasks is generalising CSS "ZX-dualities" to non-CSS codes.

Researchers found three types of Fault-Tolerant Clifford Gates operators:

Clifford physical gates and a single qubit are the only components of transversal circuits. They are Fault-Tolerant Clifford Gates because one qubit defect does not affect others. SWAP-Transversal circuits: Single-qubit Cliffords and qubit SWAP operations. The qubit architecture determines their failure tolerance. SWAP gates without qubit contact can prevent error spread in atom arrays and ion traps. SWAP gates improve the number of logical gates that can be built, as shown by the authors. General Clifford circuits: Single- and two-qubit physical gates. The new embedded code approaches limit the search to gates implementable within device connectivity limitations, but they are not guaranteed to be Fault-Tolerant Clifford Gates or low-depth. Even if the circuit distance is decreased, fault tolerance can be maintained if no qubit in a code block is used in more than one two-qubit gate.

Expanding Logical Operations

The researchers used various binary representations of stabiliser codes to select single-qubit Clifford operations for the completed circuits:

Logical operators with SWAP and Hadamard (H) gates can be found using a two-block symplectic representation ([Gx | Gz]). In combination with SWAPs, other two-block representations can identify operators like S (sqrt(Z)) or sqrt(X) gates. This is extended to any combination of single-qubit Clifford gates (H and S) and SWAPs using a new three-block model. The methods involve complex procedures to convert automorphism-group generators to physical circuits, calculate Pauli corrections to stabilise the stabiliser group's signs, and precisely determine their logical effect. The team applied embedded code to multi-qubit gates like CNOT and CZ to represent these operations as automorphisms of a larger code.

Positive Results Across Codes

New approaches were applied to several popular stabiliser codes with promising results:

The approaches show that most well-known distance codes with a single logical qubit (up to 30 physical qubits) may implement the full single-qubit Clifford group SWAP-transversally. This is a major advance over transversal gate designs and benefits quantum structures with limited qubit resources. Researchers found additional generators that re-identify grid translation automorphisms and ZX-dualities and produce a logical CNOT circuit between logical blocks in bivariate cycle codes. The collection of logical operators for these codes expands. The researchers created a Python package that uses computational algebra systems like MAGMA and the open-source Bliss package for automorphism group computations.

Future View

Code automorphisms are computationally demanding and scale exponentially with code dimension, but graph isomorphism algorithms with quasi-polynomial run-time are better at finding “matrix automorphisms” for larger codes.