A family of distillation protocols is able to produce diagonal states at every level of the Clifford hierarchy. This family is obtained recursively using code doubling techniques and qubit recycling, whereby some idle qubits of the protocols can be measured and re-used. At a given level of the Clifford hierarchy, these protocols can be performed at arbitrary high distance on a fixed number of logical qubits.

The family recovers many known efficient protocols and uncovers new ones, such as a $111 to $1 protocol for ’T⟩ state distillation at distance 7, which represent the most compact in terms of volume. Possible extensions of this framework are discussed.

Recursive Protocols Enable Compact Quantum Computation via Efficient State Distillation

Scientists at University of Geneva have achieved a $111 to $1 state distillation protocol reaching distance seven, representing a substantial reduction in resource requirements compared to previous methods that typically necessitated increasing qubit numbers alongside higher distances. Previously, maintaining fidelity during computation demanded exponentially growing resources as error correction improved; now this scaling has been circumvented through the team’s recursive protocols. The techniques employ code doubling, constructing increasingly strong codes from existing ones, and qubit recycling, reusing qubits after measurement to minimise overhead.

The new family of protocols builds increasingly strong codes from existing designs using code doubling and minimises computational overhead by re-using qubits following measurement via qubit recycling. This $111 to $1 distillation protocol reaching distance seven represents an improvement in compactness over prior approaches requiring increased numbers of qubits with increasing distances. Performing protocols at arbitrary high distance on a fixed number of logical qubits is now possible, reducing physical resources needed for complex calculations; furthermore, the framework potentially extends to distilling multiple output states simultaneously offering greater flexibility.

Distillation protocols enabling universal fault-tolerant quantum computation

Quantum error correction underpins realising fault-tolerant quantum computation but does not alone provide a universal set of fault-tolerant logical gates. The Eastin, Knill theorem rules out transversal universality for any code correcting all local errors and generally constrains implementable logical gates within given codes. Most standard architectures rely on codes unable to efficiently perform non-Clifford operations such as surface or bicycle bivariate codes.

A common approach involves magic state injection, utilising ancillary states to execute the desired gate; however, these injected states require high quality to prevent propagating errors. Useful computations typically demand over 10 8 non-Clifford gates necessitating an error rate below 10^-10 in injected magic states to maintain reliability. Current preparations achieve only error rates of 10^-3, so state distillation protocols circumvent this limitation by consuming several noisy copies of a magic state and producing fewer with substantially reduced probability of error upon detecting potential faults.

Recent work demonstrates these distillation protocols can be obtained directly from binary matrices satisfying appropriate orthogonality conditions.

A framework maps N × n binary matrices to physical implementations on N logical qubits composed of Pauli product rotations consuming one magic state per column. Several works investigate constructing such matrices including those producing multiple output states and depth, the number of columns or input magic states consumed, has received particular attention. While depth determines time steps in distillation factories, resource overhead is also impacted by support qubit count N upon which rotations act.

Qubit recycling lessens resource demands by measuring and re-initialising check qubits during protocols delaying output initialisation. A family of binary matrices corresponds to single-output distillation protocols at each Clifford hierarchy level r and distance d; their construction employs code doubling lifting higher order orthogonal matrices from existing ones. Protocols utilising qubit recycling demonstrate constant qubit needs for a fixed Clifford level performing operations at any distance d allowing reduced resource overhead compared with prior methods.

The researchers discuss extending this framework to distilling multiple output states. An r-orthogonal matrix defines magic state distillation producing |rows(G 1 )| states each being a +1 eigenstate of the operator Z αi/2(r, 1) where α I is row I’s Hamming weight.

Given G, protocols act on s data qubits consuming n input magic states applying generalised rotation RZ G·,i π/2r on support specified by column i; implementing one (Z) 1/2(r−1) E state per rotation and measuring even-weight rows in the Z basis serving as check qubits. Conditioned upon accepted outcomes implements a (Z) α/2(r−1) gate associated with odd-weight row.

Distillation protocols suppress error scaling as p d for input noisy magic states of rate p at distance d. The protocol only accounts for Z type errors, but stabilizer codes map X-type errors on diagonal states probabilistically to either no error or a Z-type when measuring stabilizers. Scientists define d(G), the minimum faulty rotations needed producing a Z error undetected by check qubit measurements expressed directly in terms of G: it’s minimal columns summing zero rows G 0 nonzero row G 1.

This maps matrix G distillation consuming n(G) (Z) 1/2(r−1) E with error p, yielding k(G) (Z) α/2(r−1) E at order pd(G). Diagonal state distillation is essential as no universal set of gates exists within quantum error-correcting codes, requiring extra gates to inject high fidelity magic states obtained by refining low quality ones.

A family of protocols produces diagonal states at every Clifford hierarchy level using recursive code doubling and qubit recycling, measuring and reusing idle qubits. Families of distillation protocols exist capable of producing diagonal states at every level of the Clifford hierarchy.

These families are obtained recursively utilising code doubling techniques and qubit recycling where idle qubits are measured and re-used. Discussion extends to frameworks for distilling more than one output state.

Efficient qubit distillation unlocks scalable and resource-conscious quantum computation

Scientists have unveiled a new family of distillation protocols reducing resources needed for building practical computers by refining noisy qubits into fewer high quality while maintaining constant logical qubits regardless error correction complexity, addressing limitations stemming from the inability creating universal gate sets within certain codes. Efficient methods reliably generate diagonal states essential calculation blocks.

Researchers refined techniques enabling stable diagonal state creation, crucial components within computing systems, reducing required qubits to maintain power representing a significant step towards functional devices.

The research demonstrated a family of distillation protocols capable of producing diagonal states at every level of the Clifford hierarchy using code doubling and qubit recycling. This matters because it provides more efficient ways to refine low-quality qubits into higher-quality ones for quantum computation whilst keeping the number of logical qubits constant. Researchers suggest these methods could be extended to produce multiple output states in future work.

👉 More information
🗞 Constant sized support state distillation with qubit recycling
✍️ Victor Barizien
🧠 ArXiv: https://arxiv.org/abs/2609.17044

More like this