SpinHex: A low-crosstalk, spin-qubit architecture based on multi-electron couplers

Rubén M. Otxoa, Josu Etxezarreta Martinez, Paul Schnabl, Normann Mertig, Charles Smith, Frederico Martins · 2025

Read SpinHex: A low-crosstalk, spin-qubit architecture based on multi-electron couplers by Rubén M. Otxoa, Josu Etxezarreta Martinez, Paul Schnabl, Normann Mertig, Charles Smith, Frederico Martins (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-03149.txt.

Introduction

Ruben Otxoa and co-authors lay spin bits on a six side grid with spin free fat dots as the links, in a paper from 2025. Spin bits in chip stone hold their state long and build with stock chip craft, a strong base for scale. Fat dots with no spin take volt calls and pass grip twixt bits at speed.

The team runs a known fault code on the grid in sim. Gate flaw gates sit at a set line past which the code wins. At near pure gates, each mind bit costs a few thousand flesh bits and a chip of palm size could host ten thousand mind bits. Counts make the grid a scale path.

Catalog record

CREATORRubén M. Otxoa, Josu Etxezarreta Martinez, Paul Schnabl, Normann Mertig, Charles Smith, Frederico Martins
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE39 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-03149

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

SpinHex: A low-crosstalk, spin-qubit architecture based on multi-electron couplers Rubén M. Otxoa, Josu Etxezarreta Martinez, Paul Schnabl, Normann Mertig, Charles Smith, Frederico Martins Abstract Semiconductor spin qubits are an attractive quantum computing platform that offers long qubit coherence times and compatibility with existing semiconductor fabrication technology for scale up. Here, we propose a spin-qubit architecture based on spinless multielectron quantum dots that act as low-crosstalk couplers between a two-dimensional arrangement of spin-qubits in a hexagonal lattice. The multielectron couplers are controlled by voltage signals, which mediate fast Heisenberg exchange and thus enable coherent multi-qubit operations. For the proposed architecture, we discuss the implementation of the rotated XZZX surface code and numerically study its performance for a circuit-level noise model. We predict a threshold of 0.18 % 0.18\% for the error rate of the entangling gates. We further evaluate the scalability of the proposed architecture and predict the need for 4480 4480 physical qubits per logical qubit with logical error rates of 10 − 12 10^{-12} considering entangling gate fidelities of 99.99 % 99.99\% , resulting in a chip size of 2.6 2.6 cm 2 to host 10,000 10,000 logical qubits. \CJKencfamily UTF8mc I Introduction Fault-tolerant quantum computers enable calculations that are impractical for classical supercomputers [ 1 , 2 , 3 , 4 ] . However, building a fault-tolerant quantum computer is difficult. For example, it requires managing hardware errors that limit the complexity of addressable computations [ 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 ] . Thus, error correction is central for realizing large-scale, fault-tolerant quantum computers. Surface codes [ 13 , 14 , 15 ] are a promising approach to realizing error-corrected quantum hardware. They allow for suppressing errors by combining many physical qubits that operate as a single logical qubit. The feasibility of this error-suppression concept has recently been demonstrated on superconducting and neutral atom hardware [ 16 , 17 , 18 , 19 ] for a single logical qubit. However, executing complex quantum algorithms requires many (logical) qubits with long coherence times [ 20 ] , which remains a significant challenge for any quantum hardware. Semiconductor spin qubits are one potential platform for realizing error-corrected quantum hardware [ 21 , 22 , 23 ] . Their benefits include small footprints, long coherence times [ 24 , 25 ] , high gate fidelity [ 26 , 27 , 28 , 29 , 30 , 31 ] , operation at 1K [ 31 ] , and the potential for scale-up using semiconductor fabrication technology [ 32 , 33 , 34 , 35 , 36 ] . Although experimental verifications of error correction protocols have just started [ 37 ] , several theoretical works propose semirealistic architectures [ 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 ] that allow for implementing surface codes with spin quibts. Unfortunately, many proposals suffer from high wiring densities and partially rely on long-range coupling methods [ 47 , 48 , 49 , 50 , 51 ] , which have not yet achieved the required fidelity. Some proposals [ 41 , 43 ] , on the other hand, significantly reduce the wiring density, by shuttling spin qubits from occupied to vacant quantum dots with high fidelity. The feasibility of this approach is under active investigation, mainly in one-dimensional (1D) arrays of quantum dots [ 52 , 53 , 54 , 55 , 56 , 57 , 58 , 57 , 58 , 59 , 60 ] . Meanwhile, experimental investigations of two-dimensional (2D) spin-qubit devices are still rare [ 61 , 62 , 63 , 59 , 64 , 65 ] . Thus, we anticipate challenges with the two-dimensional integration necessary for implementing surface codes. In particular, we expect issues due to (capacitive) crosstalk in the couplers [ 41 , 43 ] , which connect 1D arrays of quantum dots in the necessary second dimension. In this paper, we propose SpinHex – a semirealistic device design for operating surface codes with spin qubits at low crosstalk. SpinHex provides a 2D arrangement of spin qubits in a hexagonal lattice. Its core unit is a spinless multi-electron quantum dot, which mediates voltage-controlled Heisenberg exchange [ 66 , 67 , 68 ] between up to three spin qubits in 2D. Such a multi-electron coupler (MEC) provides a large separation between the coupled spin qubits, thus significantly lowering capacitive crosstalk. To evaluate the feasibility of SpinHex, we discuss the implementation of the rotated XZZX surface code [ 14 ] and numerically study its performance under device-specific circuit-level noise. In particular, we consider the errors introduced by swapping spin-qubits across MECs and report a threshold of 0.18% on the error rate of the entangling gates. We also compute the number of physical qubits required to reach the MegaQuop , GigaQuop , and TeraQuop regimes, where a single logical qubit can perform a million, a billion, or a trillion error-free quantum operations, respectively. At error rates of 0.01% we show that approximately 1,000 physical qubits are required to reach the MegaQuop regime. Meanwhile, a footprint of approximately 4480 qubits is required to reach the TeraQuop regime. Finally, we estimate that a chip operating 10,000 logical qubits in the TeraQuop regime would result in a chip of around 2.6\mathrm{cm}^{2} . Figure 1: (a) SpinHex hardware graph composed of MECs (red discs) and 1D quantum dot arrays (black lines). Cirles (yellow and blue) represent the data and measurement qubits of a logical qubit (blue background tiles). (b) MEC coupling the occupied quantum dots (blue disc) of three 1D quantum dot arrays. (c-e) Three possible realizations of a 1D quantum dot array. Blue circles represent empty quantum dots. The underlined quantum dots host the surface code qubit. II Results II.1 SpinHex hardware graph The hardware graph of SpinHex is depicted in Fig. 1 . Similar to [ 43 ] , SpinHex connects 1D arrays of quantum dots via three-way junctions to form a hexagonal lattice of data and measurement qubits, see Fig. 1 (a). Instead of T-junctions [ 43 ] , we propose using spinless multi-electron (or hole) quantum dots [ 66 , 67 , 68 ] to couple the terminal ports of three 1D quantum dot arrays in 2D, see Fig. 1 (b). Such multi-electron (or hole) couplers (MECs) provide a large separation between the terminal quantum dots of the coupled 1D quantum dot arrays, resulting in low capacitive crosstalk. Finally, we envision three possible realizations of the 1D quantum dor arrays, as depicted in Figs. 1 (c-e): In the first realization, Fig. 1 (e), the 1D array consists of empty quantum dots. Two occupied quantum dots host a compute and reference spin-qubit for readout at the array’s center. The spin-qubit is shuttled along the 1D array to SWAP its information with other 1D arrays across the MECs at the terminal. This realization is similar to replacing T-junctions in Ref. [ 43 ] with three-way MECs. Due to the high fidelity of shuttling operations, this is likely an interesting realization. In the second realization, Fig. 1 (d), the quantum dots of the 1D array are all occupied by one spin-qubit. Two quantum dots host a

Read the complete text