A plug-and-play superconducting quantum controller at millikelvin temperatures enables exceeding 99.9% average gate fidelity
Kuang Liu, Zhiyuan Wang, Xiaoliang He, Siqi Li, Hao Wu, Xiangyu Ren, Zhengqi Niu, Wangpeng Gao
Abstract
The development of large-scale superconducting quantum computing requires efficient in-situ control methods that allow high-fidelity operations at millikelvin temperatures. Superconducting circuits based on Josephson junctions offer a promising solution due to their high speed, low power dissipation, and cryogenic nature. Here, we report a superconducting quantum controller that enables direct chip-to-chip interconnection with qubits at 10 mK and high-fidelity, all-digital manipulation. Randomized benchmarking reveals a uniformly high average Clifford fidelity of 99.9% with leakage to high energy levels on the order of , and an estimated average gate operation energy of 0.121 fJ, demonstrating the potential to resolve the control bottleneck in superconducting quantum computing.
AI Impact Assessments
(3 models)Scientific Impact Assessment
Core Contribution
This paper presents a superconducting quantum controller based on Single Flux Quantum (SFQ) pulse technology that operates at 10 mK in a plug-and-play configuration with direct chip-to-chip interconnection to transmon qubits. The key claim is achieving 99.90(1)% average Clifford gate fidelity via randomized benchmarking—a significant improvement over prior SFQ-based control demonstrations. The controller employs several design innovations: a fully passive superconducting bias network to eliminate static-power-related quasiparticle generation, a low-critical-current topology to minimize dynamically generated quasiparticles, and an on-chip spectrum-engineering unit to filter out-of-band harmonics that cause leakage to higher energy levels. The average gate operation energy is estimated at 0.121 fJ, orders of magnitude below room-temperature electronics.
Methodological Rigor
The experimental characterization is reasonably thorough. The authors demonstrate:
1. Rabi and Ramsey experiments confirming coherent control with clear chevron patterns and interference fringes.
2. Standard and interleaved randomized benchmarking (RB/IRB) yielding consistent fidelity metrics across the full 24-Clifford gate set.
3. Purity RB showing the decoherence-limited error (8.91×10⁻⁴ per Clifford) is close to the standard RB error, indicating performance near the coherence limit.
4. Leakage RB comparing SFQ-driven leakage (2.90×10⁻⁴) against conventional microwave-driven leakage (1.46×10⁻⁴), showing the SFQ approach is comparable.
5. Thermal excitation measurements using JPA-assisted single-shot readout to verify that controller operation does not measurably increase qubit excited-state population.
However, several limitations in rigor should be noted. The work demonstrates only single-qubit gates on what appears to be a single qubit. No two-qubit gate demonstrations are provided—these are deferred to future work. The Clifford set relies heavily on virtual-Z gates (which are inherently ideal), meaning only a subset of gates involve physical SFQ pulses. While this is a standard technique, it means the "uniformly high fidelity" claim partially benefits from the inclusion of zero-error virtual gates. The paper also lacks detailed error budget decomposition beyond purity and leakage analysis, and simultaneous benchmarking results (e.g., gate set tomography) that could reveal coherent error structure are absent.
The comparison to the "publicly reported record" (Ref. [24], Jordan et al. 2026 in Nature Electronics) is made but without detailed side-by-side analysis of experimental conditions, qubit coherence times, or controller architectures that would allow fair comparison.
Potential Impact
This work addresses a genuine and critical bottleneck in superconducting quantum computing: the wiring and control signal delivery problem. Current systems require extensive room-temperature electronics connected through complex cryogenic wiring harnesses, and this approach fundamentally limits scalability beyond ~1000 qubits in existing dilution refrigerators. A cryogenic controller that achieves error rates below the surface code threshold (~1%) while consuming femtojoules of energy per gate represents meaningful progress toward resolving this bottleneck.
The plug-and-play aspect—discrete chip architecture connected via standard coaxial cables—is practically significant because it allows flexible deployment and doesn't require monolithic integration with the qubit chip, which would impose fabrication constraints and potential crosstalk issues.
If these results extend to multi-qubit systems with two-qubit gates, the impact would be substantial. However, the gap between single-qubit demonstrations and practical multi-qubit control systems remains large, involving challenges in multiplexing, crosstalk management, and simultaneous control of many qubits.
Timeliness & Relevance
This paper is highly timely. The quantum computing community is actively grappling with the "wiring problem" as processors scale beyond 100 qubits. Multiple groups (McDermott/Plourde at Wisconsin, NIST/Boulder, SeeQC/Oxford) are pursuing SFQ-based control, and the race to demonstrate high-fidelity cryogenic control is intensifying. The concurrent Nature Electronics publication by Jordan et al. (Ref. [24]) underscores the competitive landscape. Achieving 99.9% fidelity places this work at the leading edge and makes it directly relevant to near-term error correction experiments.
Strengths
Limitations
Overall Assessment
This is a strong experimental demonstration that advances the state of the art in cryogenic quantum control. The 99.9% fidelity milestone is significant and timely. However, the path from single-qubit control to a scalable multi-qubit control system involves substantial additional challenges that remain unaddressed. The paper makes a clear incremental but important contribution to a critical problem in quantum computing.
Generated Apr 8, 2026
Comparison History (38)
Paper 2 dramatically lowers the physical qubit requirement for breaking RSA-2048 from millions to under 100,000. This breakthrough in fault-tolerant architecture using QLDPC codes profoundly impacts the timeline for utility-scale quantum computing and accelerates the urgent global need for post-quantum cryptography, giving it extraordinary real-world relevance and cross-disciplinary impact.
Paper 2 presents a fundamentally new theoretical framework ('quantum uploading' and 'Heisenberg learning tree' method) proving exponential speedups in fault-tolerant quantum learning from experiments, with broad implications across quantum computing theory, quantum sensing, and experimental physics (e.g., astronomical imaging). Its results establish new separation theorems and introduce general-purpose proof techniques. Paper 1, while impressive engineering (99.9% fidelity cryogenic controller), represents an incremental advance in quantum control hardware. Paper 2's breadth of impact, theoretical novelty, and cross-disciplinary applicability give it higher potential scientific impact.
Paper 1 demonstrates a superconducting quantum controller operating at millikelvin temperatures achieving 99.9% average gate fidelity with ultra-low energy dissipation (0.121 fJ). This addresses the critical scalability bottleneck of control wiring in quantum computing—a fundamental challenge for large-scale systems. The plug-and-play chip-to-chip integration at cryogenic temperatures represents a paradigm shift in quantum control architecture. While Paper 2 presents a clever adaptive protocol for TLS-induced bistability, it addresses a more specific error source. Paper 1's broader architectural implications for scaling quantum processors give it higher potential impact.
Paper 1 likely has higher impact: it addresses a core scalability bottleneck for fault-tolerant quantum computing across hardware platforms by delivering large, quantified improvements in decoding accuracy, latency, and throughput for modern LDPC codes, and introduces a new “waterfall” operating regime with implications for overall space-time cost. Its contributions are broadly applicable (QEC theory, ML for decoding, systems architecture) and timely as LDPC fault tolerance gains prominence. Paper 2 is strong and practical for superconducting systems, but is more platform-specific and incremental relative to the larger, cross-cutting leverage of improved decoders.
Paper 2 likely has higher impact because it bridges a major architecture gap between physical hardware constraints and QEC/algorithm requirements, providing an end-to-end heterogeneous architecture, compiler, and detailed resource accounting at large logical scales. The reported 138× physical-qubit reduction and concrete RSA-2048 projections are broadly relevant to fault-tolerant quantum computing roadmaps, influencing hardware, compiler, and QEC communities. While Paper 1 is a strong experimental advance for cryogenic control, its impact is more specialized to superconducting platforms and depends on integration at scale; Paper 2’s framework can affect multiple modalities and near-term planning.
Paper 1 addresses a critical hardware bottleneck in scaling superconducting quantum computers by demonstrating a functional millikelvin controller with extremely high gate fidelity and low power. This experimental breakthrough directly enables the near-term scaling of quantum devices, offering high real-world applicability. While Paper 2 presents a significant theoretical protocol for quantum networks, Paper 1's concrete hardware advancement is likely to have a more immediate and widespread impact on the practical development of large-scale quantum systems.
Paper 1 addresses a critical scalability bottleneck in quantum computing—control electronics at millikelvin temperatures—achieving 99.9% gate fidelity with extremely low energy dissipation. This directly enables scaling to large-scale quantum computers, a grand challenge with enormous breadth of impact. Paper 2 makes important advances in quantum-limited communications with integrated coherent receivers, but its impact is more narrowly focused on optical communications. The quantum computing scalability problem addressed by Paper 1 has broader transformative potential across multiple scientific and technological domains.
Paper 2 likely has higher scientific impact due to its immediate, scalable real-world application: solving the cryogenic control bottleneck for superconducting quantum computers while demonstrating >99.9% average gate fidelity and ultra-low energy operation. The combination of a plug-and-play millikelvin controller, chip-to-chip interconnect, and strong benchmarking suggests high methodological rigor and near-term adoption across quantum hardware efforts. Paper 1 is novel and valuable for hybrid CV-DV QEC without GKP states, but its impact may be more specialized and dependent on experimental feasibility across platforms.
Paper 1 presents a crucial experimental breakthrough in quantum computing hardware, addressing the critical scalability bottleneck of cryogenic qubit control. Achieving >99.9% fidelity with ultra-low power at 10mK has immediate, profound real-world applications for building large-scale quantum computers. While Paper 2 offers deep theoretical insights into gauge symmetry and error correction, Paper 1's tangible engineering achievement will likely drive broader and more immediate technological impact across the quantum computing industry.
Paper 1 demonstrates a practical superconducting quantum controller operating at millikelvin temperatures achieving 99.9% gate fidelity with ultra-low energy dissipation (0.121 fJ). This addresses a critical scalability bottleneck—the wiring and control electronics problem—for large-scale quantum computing. Its experimental demonstration of plug-and-play chip-to-chip control at 10 mK has immediate engineering impact and broad relevance across the entire superconducting quantum computing community. Paper 2 makes an important theoretical contribution to magic state injection for qLDPC codes, but addresses a more specialized problem with primarily simulation-based results and narrower near-term applicability.