ETH Zurich publishes peer-reviewed Nature paper on experimental randomness amplification using entangled superconducting qubits over a 30-meter cryogenic link, certifying perfect-quality random output from imperfect inputs
On 2026-05-27 (Wednesday), researchers at ETH Zurich's Department of Physics — led by experimentalist Andreas Wallraff (Quantum Device Lab) and theorist Renato Renner (Institute for Theoretical Physics) — published 'Experimental Randomness Amplification' in Nature (DOI: 10.1038/s41586-026-10521-8; full author list: Kulikov A, Storz S, Schär JD, Sandfuchs M, Wolf R, Berterottière F, Hellings C, Wallraff A, Renner R). The experiment uses the team's previously-demonstrated 30-meter cryogenic-link superconducting-qubit platform (the same platform used for the 2023 loophole-free Bell-inequality violation reported in Nature 617, 265-270) to perform a Bell-test-certified randomness-amplification protocol. The protocol takes imperfect random inputs (e.g. from a classical noise source with biased entropy) and produces a sequence of output bits whose randomness is mathematically certified using the observed Bell-inequality violation — converting partially-random inputs into 'perfectly random' outputs whose certification does not depend on the trustworthiness of the underlying hardware (a device-independent randomness amplification scheme). The ETH Zurich press release issued the same day characterizes the result as a milestone in the field, with potential applications across secure-key generation, blockchain consensus, lottery systems, and other cryptographic infrastructure requiring certified randomness. The press release does not disclose specific named-metric throughput (random bits per second), Bell-test S-value, or certification confidence-level numbers; reliance on the full Nature paper is required for those details. The 30-meter cryogenic link itself is a notable superconducting-quantum infrastructure asset — to the system's knowledge it remains the longest-distance loophole-free Bell test on a superconducting platform, with the previously-reported S value of 2.0747±0.0033 violating Bell's inequality at P<10^-108 confidence across ~1M experimental trials.
Score 6 — anchor §8.2 row 6 'Credible benchmark result with industry-wide implications (e.g. T1 coherence record, boson-sampling scale, sensor sensitivity record). Threshold for when a research item becomes investor-relevant.' Held at 6 not 7 because (a) the certification protocol is a research demonstration not a deployable product; (b) the press release lacks named-metric throughput numbers (bits-per-second random output, certification confidence interval), with operational reliance on the Nature paper for verification; (c) no commercial deployment partner is named at announcement. Held above 5 because (a) peer-reviewed Nature publication is the highest-tier verification venue; (b) Renner and Wallraff are first-tier PIs (Renner is a foundational quantum-information-theory contributor; Wallraff's Quantum Device Lab is one of the most cited circuit-QED experimental groups globally); (c) the underlying 30-meter cryogenic-link platform is itself an industrially-significant superconducting-quantum infrastructure asset; (d) device-independent randomness amplification has direct CISO-audience relevance for PQC-era cryptographic primitive generation — randomness is the upstream primitive that all post-quantum key-encapsulation and signature schemes depend on for security. Per §8.2 'Trap 7 — Under-rating standards. Policy and standards events feel dry, but PQC standards activity is load-bearing for the CISO audience.' Device-independent certified randomness is structurally analogous to a PQC primitive in this sense — it is the upstream foundation that enables higher-level cryptographic guarantees.
Three trackable second-order consequences. First, the device-independent randomness-amplification protocol's commercial productization path runs through quantum-random-number-generator (QRNG) vendors (ID Quantique, QuintessenceLabs, Quside) plus HSM vendors (Thales, Entrust, Utimaco) — watch for which vendor first claims a Nature-paper-derived certified-randomness reference implementation; the ETH Zurich licensing/IP terms (not disclosed in the press release) determine whether commercial uptake is open-source-style or requires per-implementation license fees. Second, the 30-meter cryogenic-link platform's demonstration-record status — Bell-test 2023 + randomness-amplification 2026 — positions ETH Zurich as a credible third-party validation venue for upcoming superconducting-quantum-link claims from competitors (Yale Quantum Circuits dual-rail, IBM modular superconducting interconnect roadmap, Rigetti modular chiplet); watch for cross-validation collaboration announcements between ETH and named commercial superconducting vendors. Third, the certified-randomness primitive interacts with the broader 2026-Q2 PQC-migration policy stack — G7 QTWG financial-sector framework (2026-05-11), NIST Additional Digital Signatures Round 3 (2026-05-14), GlobalPlatform Pavona open-silicon ML-KEM/ML-DSA distribution (2026-05-26): the cryptographic-primitive supply chain now spans key-encapsulation (ML-KEM), digital-signature (ML-DSA + 9 Round 3 candidates), and root randomness (this ETH paper's path); if any major HSM vendor adopts a Nature-paper-derived certified-randomness primitive within the next 12 months, that closes a previously-implicit gap in the post-quantum cryptography hardware-implementation stack.