The New Quantum Era - innovation in quantum computing, science and technology

Dr. Shintaro Sato is a Fellow and Head of the Quantum Laboratory at Fujitsu Research, and Deputy Director of the RIKEN RQC-Fujitsu Collaboration Centre. He oversees Fujitsu's entire quantum effort — from device fabrication through error correction architecture, software, and application research — and has been building toward commercial quantum systems since Fujitsu began its serious quantum R&D push around 2020. With over 164 publications spanning graphene nanoelectronics, superconducting qubit design, and fault-tolerant architectures, he brings both deep technical credibility and a rare full-stack perspective.This conversation is timely because two significant developments converged almost simultaneously just before recording: Fujitsu announced a tin-vacancy (SnV) diamond-spin prototype developed with TU Delft and QuTech, and began testing its STAR error-correction architecture on neutral-atom hardware with startup Yaqumo — an explicit signal that Fujitsu is betting on hardware-agnostic software layers even as it races to scale its own superconducting devices. Listeners who follow quantum hardware roadmaps, fault-tolerant computing, or Japan's national quantum strategy will find this episode unusually specific and candid.What We Get IntoHow Fujitsu actually fabricates its superconducting chips — Sato explains how the team adapted Nakamura-sensei's original RIKEN designs, developed their own Josephson junction uniformity techniques, and built a research fab capability from scratch rather than licensing finished devices.Why packaging is the hardest problem at 1,000+ qubits — not the qubits themselves, but the superconducting wiring, chip-to-chip interconnects, interposers, and the sheer number of control lines running from room temperature to millikelvin. Sato is candid that his engineers "don't want to do it anymore."What STAR architecture actually does — it's not a new error-correction code; it operates at the logical gate layer above the surface code, replacing the notoriously expensive T gate with phase-rotation gates and selectively reintroducing T gates only when rotation angles would otherwise accumulate too much error. Version 3 achieves roughly a 10x accuracy improvement over version 2 for the same physical qubit count.Why STAR is hardware-agnostic — because it operates at the logical operation layer rather than the physical error-correction layer, it can in principle run on any hardware modality. Fujitsu has now begun testing it on Yaqumo's neutral-atom platform, and Sato mentions QuEra has also demonstrated STAR independently.What tin-vacancy (SnV) diamond-spin qubits are actually for — not a replacement for superconducting qubits, but a photonic interconnect technology. SnV centers emit photons that can carry quantum information between modules, potentially linking separate dilution refrigerators — a transduction approach that sidesteps one of the hardest problems in scaling superconducting systems.The speed mismatch problem in hybrid architectures — superconducting qubits operate at gigahertz speeds; nuclear spins in diamond-spin systems operate at kilohertz. Sato acknowledges this directly and frames it as an architectural challenge analogous to CPU-memory hierarchies in classical computing.What Fujitsu's open-source strategy looks like in practice — the collaboration with Osaka University on Project Octopus, which produced an open-source full-stack control and software platform, is now being adopted as part of Fujitsu's commercial offering.What 2030 actually looks like — Sato is measured: a 10,000-qubit machine will still be a "very small scale logical quantum computer," most useful for quantum chemistry in hybrid HPC+quantum workflows. Revolutionary applications will come incrementally, not all at once.The long-term vision — Sato's stated dream is a quantum computer small enough to fit in a smartphone, which he acknowledges requires entirely different physics than anything on today's roadmap.Resources & LinksGuest & OrganizationDr. Shintaro Sato — Fujitsu Quantum Day Profile — Fujitsu's own bio page covering Sato's dual role at Fujitsu Research and RIKEN RQC-Fujitsu Collaboration Centre.Shintaro Sato — ResearchGate Profile — Full publication list (164 papers, 3,096 citations); useful for tracing his path from graphene nanoelectronics to quantum architecture.Interview with Shintaro Sato — QuTech — Sato discusses the NV-center and diamond-spin collaboration with QuTech and the vision for quantum-as-a-service.Hardware Roadmap & STAR ArchitectureFujitsu Quantum Day 2026 — Sato Presentation Slides (PDF) — Primary-source slides showing the full hardware roadmap: 64 → 256 → 1,024 → 10,000+ qubits with STAR architecture milestones.Fujitsu Officially Starts 10,000+ Qubit Development — Fujitsu Press Release (Aug 2025) — Official announcement of the early-FTQC 10,000-qubit program targeting FY2030.STAR Architecture Overview — Fujitsu Global — Technical explainer for STAR ver. 3 (announced March 2026), including the qubit-reduction claims for ruthenium-catalyst calculations.How Fujitsu Is Tackling a 10,000-Qubit Quantum Computer — The Quantum Insider (Dec 2025) — Deep-dive on STAR architecture and the 250-logical-qubit target.Diamond-Spin & SnV PrototypeFujitsu Develops Diamond-Spin Quantum Computer Prototype — The Quantum Insider (Sept 2026) — Announcement of the SnV-center diamond-spin QPU prototype from the Fujitsu/TU Delft/QuTech collaboration.Fujitsu Diamond-Spin SnV Prototype Analysis — Quantum Computing Report (Sept 2026) — Technical breakdown of SnV center advantages, photon emission characteristics, and hybrid roadmap context.Fujitsu and QuTech Realize High-Precision Quantum Gates — Fujitsu Press Release (Mar 2025) — NV-center two-qubit gate fidelity result (<0.1% error) published in Physical Review Applied; the technical foundation for the SnV work discussed in this episode.Neutral-Atom / Yaqumo CollaborationFujitsu and Yaqumo Begin Testing on Neutral-Atom Hardware — The Quantum Insider (Sept 2026) — Breaking news on STAR validation on Yaqumo's neutral-atom processors, announced just before this recording.

Marie Lepske brings a combination that's genuinely rare in venture: a background in applied mathematics and physics, early experience covering quantum at Runa Capital before most generalist funds knew the field existed, and now a GP seat at Constructor Capital, which closed a $110M Fund I in February 2026 with more than half its capital directed toward next-generation computing including quantum. She backed Qnami at Runa — a quantum sensing company acquired by Quantum Design in June 2026, one of the few clean sensing exits the field has produced — and Constructor's portfolio includes QuEra, which raised over $230M in a round led by Google Quantum AI and SoftBank.The conversation matters now because the quantum investment landscape is genuinely changing. SPAC activity, mega-rounds from hyperscalers, and rising valuations are pulling in non-specialist capital at the same time that the science is getting harder to evaluate from the outside. Lepske is one of the people who has to navigate that tension every day, and she's willing to name the failure modes.Founders building quantum or deep-tech companies, investors trying to understand where specialist and generalist capital intersect, and technically curious listeners who want to understand how the business side of quantum actually works will all find this episode useful.What We Get IntoWhy technical training matters at the earliest stages — and what it actually buys you when a founding team's only asset is a laboratory and an optical table, before there's even a legal entityHow the specialist-versus-generalist divide is evolving — Lepske's 2022 argument that early-stage quantum would stay specialist territory, and how she reads the arrival of Google, SoftBank, and NVIDIA in QuEra's cap tableThe due diligence framework for pre-product science startups — team provenance, patent landscape, IP legal review, and how Constructor uses its scientific advisory board to validate founders they can't fully assess internallyWhat a "no" looks like — the specific signals that make Constructor wait rather than invest, including teams with strong science but no credible path to market and IP positions that are already crowdedThe Qnami exit and what it reveals about quantum sensing — why sensing applications are currently concentrated in R&D markets, which verticals Lepske thinks will break out first, and why sensing competes with computing for buyer attentionThe SPAC problem — why the wave of quantum public listings concerns her, and why evaluating roadmap credibility requires the kind of deep familiarity that takes years to buildHow Constructor supports portfolio companies beyond capital — executive hiring, software development guidance, co-investor introductions, and the limits of that hands-on modelAI and quantum as parallel tracks — why she doesn't see AI as cannibalizing quantum talent or capital, and where she thinks the two fields will eventually convergeResources & LinksGuest & FundMarie Lepske — LinkedIn — Primary professional profile; includes keynote highlights and 2026 speaking engagementsConstructor Capital — Official Website — Fund homepage with portfolio listings, team bios, and investment thesis across DeepTech, Software Tech, and Knowledge TechConstructor Capital Fund I Close (Tech.eu, Feb 2026) — Coverage of the $110M close, check sizes ($1–10M, select up to $15M), and the university sourcing networkPortfolio Companies ReferencedQuEra — $230M+ Financing Round Announcement — The round led by Google Quantum AI and SoftBank Vision Fund 2, with NVIDIA's NVentures participatingQnami Acquisition by Quantum Design (The Quantum Insider, June 2026) — The sensing exit discussed in the episode; context on what a quantum sensing acquisition looks likeBackground ReadingQuantum Computing Report — "Venture Capital Trends in Quantum Technologies" (Runa Capital, 2022) — Lepske's co-authored market analysis, the source of her "specialist VC" thesis discussed in the episodeConstructor Capital — "Quantum Investing Playbook" (Nov 2025) — Her Paris keynote on making successful quantum investments; source of the valuation figures cited in the conversationEntangled Future — Quantum Computing Funding & Investment Landscape 2026 — Context on the $4.9B private VC figure for 2025 and the broader public commitment landscapeRecent Constructor NewsConstructor Start Demo Day Finalists (The Quantum Insider, Sept 2026) — Quantum startups among the 16 finalists announced one week before this recordingKey Quotes & InsightsOn what early-stage diligence actually looks like: > "They have only laboratory, and they have optical table which they want to show to you and you should understand if they have some interesting and useful patents or they can file them during the next one to two years."On where specialist and generalist capital divide: > "These bigger players, they're coming later. They're coming when we already investigated that this particular technology and this particular team actually could win."On the SPAC problem: > "It's very difficult for [non-specialist investors] to evaluate if they're doing proper investment or not, or if this valuation is good or not… You should be able to find out these roadmaps and to understand if this is a real roadmap or just written for the SPAC."Insight — sensing vs. computing for buyer attention: Lepske observes that quantum sensing is structurally underappreciated relative to computing, not because the products aren't real, but because governments and corporates perceive computing as the larger prize — which she suggests is not obviously correct.Insight — quantum talent scarcity as a leading indicator: She notes that quantum computing companies are now competing for talent the way AI companies did four or five years ago, with roughly 400–450 relevant laboratories globally — a number that puts the field's scale in sharp relief.Related EpisodesEp. 22 — Trapped Ions and Quantum VCs with Chiara Decaroli — Another physicist-turned-investor conversation; useful companion for understanding how scientific training shapes investment judgment in quantum hardwareEp. 96 —

Tim Palmer is not a quantum computing skeptic from the outside. He is a Fellow of the Royal Society, a CBE, an IPCC lead author, and the inventor of probabilistic ensemble forecasting — techniques now used in every major weather prediction center on Earth. He did his PhD in general relativity under Roger Penrose. When someone with that profile publishes a peer-reviewed paper in PNAS arguing that the entire fault-tolerant quantum computing roadmap may rest on a mathematical assumption that is subtly and profoundly wrong, it is worth paying close attention. The timing matters. The quantum computing industry is spending billions on the assumption that standard quantum mechanics scales indefinitely — that if you can build enough error-corrected qubits, Shor's algorithm will eventually factor RSA-2048. Palmer's RaQM framework, reviewed by leading quantum foundations researchers including Lucien Hardy and Nicolas Gisin, makes a concrete, falsifiable prediction that this assumption will fail somewhere between 200 and 1,000 error-corrected qubits. That falsification window is opening right now. This episode is for anyone who cares about the foundations of quantum mechanics, the long-term viability of fault-tolerant quantum computing, or the rare pleasure of watching a serious scientist put a real stake in the ground. What We Get Into Why Palmer left general relativity and came back to quantum foundations decades later — the conceptual thread connecting chaos theory, climate forecasting, and the geometry of quantum state space What Rational Quantum Mechanics actually proposes — specifically, why Palmer argues that Hilbert space should be defined over rational numbers rather than the full continuum of complex numbers, and what that means physically The information-theoretic argument for a qubit ceiling — why, in RaQM, the information content of n entangled qubits grows linearly with n rather than exponentially, and why that creates a hard conflict with the quantum Fourier transform above a few hundred qubits Why the 2022 Nobel Prize did not prove non-locality — Palmer's careful distinction between "Bell's inequality is violated" (experimentally established) and "the world is non-local" (an interpretation), and why that distinction matters enormously How gravity enters the picture — Palmer's argument, drawing on Penrose-style thinking, that gravity is the physical mechanism responsible for discretizing Hilbert space, and how that determines the numerical qubit ceiling Why the ceiling is technology-dependent but bounded absolutely — how Palmer estimates ~400 qubits for current photonic technology and argues that no technology, however exotic, can push the limit above ~1,000 What happens to Shor's algorithm specifically — why the quantum Fourier transform is the precise point of failure, and what that means for RSA encryption and the post-quantum cryptography transition The optimistic read on a potentially negative result — Palmer's argument that a fundamental discovery about quantum mechanics, even one that limits quantum computing, could open doors we cannot yet imagine, in the same way general relativity eventually gave us GPS Resources & Links Guest Tim Palmer — Oxford Department of Physics — Full publication list, including the RaQM paper and prior quantum foundations work Royal Society Fellow Profile: Professor Tim Palmer CBE FRS — Official Fellow profile covering Palmer's career from ensemble forecasting to quantum foundations Oxford Quantum Institute: Tim Palmer — Confirms Palmer's affiliation with OQI Papers & Articles "Rational Quantum Mechanics: Testing Quantum Theory with Quantum Computers" — PNAS (March 2026) — The primary paper discussed in this episode; the full RaQM proposal with the qubit-ceiling prediction arXiv preprint: "Rational Quantum Mechanics: Testing Quantum Theory with Quantum Computers" (Feb 2026) — Full preprint version with acknowledgements and references arXiv preprint: "Solving the Mysteries of Quantum Mechanics: Why Nature Abhors a Continuum" (Feb 2026) — Companion paper laying out the philosophical and mathematical case for RaQM; good starting point for the conceptual argument Oxford Physics News: "Rational Quantum Mechanics — A New Theory of Quantum Physics" (March 17, 2026) — Oxford's official press release; accessible summary of RaQM and its implications IAI TV: "New Theory Argues Quantum Physics Must Abandon Irrational Numbers and the Continuum" (May 2026) — Palmer's own popular-audience explanation of RaQM; a good read before or after the episode IAI TV: "Chaos Theory Eliminates Quantum Uncertainty" — Palmer's accessible argument that quantum uncertainty is epistemic rather than fundamental The Quantum Insider: "Is RSA Safe? New Study Argues Quantum Computers Face a Hard Ceiling" (March 2026) — Industry-facing coverage of RaQM's implications for cryptography PostQuantum.com: "The 1,000-Qubit Ceiling That Probably Isn't" (April 2026) — Skeptical technical rebuttal of RaQM's qubit-ceiling prediction; worth reading alongside the paper ECMWF Festschrift: "Tim's Adventures in Quantum Mechanics" (April 2026) — Palmer's own account of his transition from meteorology to quantum foundations; illuminating backstory arXiv: "Quantum Computers for Weather and Climate Prediction: The Good, the Bad and the Noisy" (Tennie & Palmer, 2022) — Palmer's earlier, more optimistic assessment of quantum computing for climate modeling; an interesting counterpoint to RaQM's ceiling prediction Further Reading The Primacy of Doubt — Oxford University Press — Palmer's 2022 trade book synthesizing chaos, climate, and quantum uncertainty; accessible entry point to his broader thinking 214 is a delightful bonus to the conversation, a short story by Noel Gorelick that imagines what might happen if RaQM turns out to be true. Key Quotes & Insights > "It's not that nature ...

Daniel Loss is RDIA Chair Professor of Quantum Computing and Director of the Quantum Center at King Fahd University of Petroleum and Minerals in Saudi Arabia, where this work was done. He is also one of the most influential theorists in quantum computing. The 1997 Loss-DiVincenzo proposal — that electron spins in quantum dots could serve as qubits — now has more than 9,000 citations and is the conceptual foundation for the semiconductor spin-qubit platforms that Intel, HRL, Diraq, and a wave of European startups are actively building toward. In 2025, Loss was named a Clarivate Citation Laureate in Physics, a designation with a strong historical track record as a Nobel Prize predictor. The reason to listen now is that Loss has turned his attention to a question that predates quantum computing itself: can reversible, energy-efficient classical logic be physically realized? His 2026 paper argues that the spin-qubit hardware the field has spent decades developing is, almost incidentally, the ideal platform to do exactly that — and that the energy advantage over room-temperature CMOS could be so large it would matter enormously for AI inference workloads and data-center power budgets. This episode is for anyone following the spin-qubit roadmap, the energy crisis in classical computing, or the deeper question of what semiconductor quantum hardware is ultimately good for. What We Get Into Why reversible computing is having a moment now: The ideas of Landauer, Bennett, Fredkin, and Toffoli have been around since the early 1980s — Loss explains what changed experimentally that makes the proposal feel like engineering rather than philosophy. The core energy claim, unpacked: Loss walks through why a spin-qubit Toffoli gate operating near 4 Kelvin could cost roughly 10⁵ times less energy than its CMOS equivalent, even after accounting for refrigeration overhead — and where the accounting is still incomplete. What makes the computation classical and the hardware quantum: The gate uses coherent quantum dynamics internally, but inputs and outputs are classical spin states. No superposition is required between gate operations, which dramatically relaxes the error-correction burden. The iToffoli gate and why it works for classical logic: Loss describes how a target spin hopping between quantum dots, controlled by two neighboring spins, implements a universal reversible gate using only DC voltage pulses — no radio-frequency drives required. The Quantum Zeno trick for classical memory: Frequent projective measurement of a spin state can stabilize it against relaxation, turning one of quantum computing's central headaches into a feature for classical storage. Why AI inference is the natural first application: Error tolerance, parallelizability, and the absence of a need for new algorithms make AI inference workloads a compelling early target — and Loss argues this de-risks the spin-qubit enterprise regardless of whether fault-tolerant quantum computing arrives on schedule. The dual-use platform argument: The same germanium/silicon quantum-dot array could, in principle, run quantum algorithms when superposition is useful and classical reversible logic when it is not — a flexibility that changes the economic calculus for building the hardware. What the experimental community needs to do next: Loss describes the first falsifiable tests — a three-spin iToffoli truth table, energy budget measurements, and a five-dot reversible adder — and names the groups best positioned to run them. How the brain comparison reframes the stakes: Loss notes that his energy estimates put spin-qubit classical computing roughly four orders of magnitude below the energy cost of a biological synapse, which has implications for how we think about the physical limits of AI. Resources & Links Guest & Lab Daniel Loss' profile at King Fahd University of Petroleum and Minerals Daniel Loss — University of Basel, Condensed Matter Theory & Quantum Computing Group — Loss's lab page; lists current research areas, group members, and leadership roles including NCCR SPIN. Daniel Loss — Wikipedia — Comprehensive biographical overview, career history, and awards list for listeners who want background before or after listening. Papers & Articles Classical Reversible Computation by Quantum Coherence — arXiv:2607.06219v3 (2026) — The central paper discussed in this episode; v3 adds a reversible-adder blueprint, control-electronics energy analysis, and optimized pulse sequences. Loss-DiVincenzo: Quantum Computation with Quantum Dots — Physical Review A (1998) — The foundational 1997/1998 proposal that started the spin-qubit field; cited more than 9,000 times and the direct ancestor of the hardware Loss now proposes for classical reversible computing. Long-Range Crossed Andreev Reflection in a Topological Insulator Nanowire — Nature Physics (2025) — A 2025 result from the Loss group with implications for topological quantum computing, showing the breadth of the research program surrounding this episode's topic. Prof. Daniel Loss Named Citation Laureate 2025 in Physics — NCCR SPIN — Announcement of Loss's Clarivate recognition; useful context for why this conversation is happening now. Organizations NCCR SPIN — National Center of Competence in Research: Spin Qubits in Silicon — Switzerland's national spin-qubit research center, co-directed by Loss; the institutional home of much of the experimental work Loss references. Max Planck Institute of Microstructure Physics — Loss's new external scientific membership, signaling a push to connect topological quantum magnetism theory with experiment. Key Quotes & Insights > "If the physical platform is successful, we are guaranteed to have killer applications — because I don't need to find new algorithms for this." > — Loss on why classical reversible computing de-risks the spin-qubit investment, regardless of the timeline for fault-tolerant quantum algorithms. > "The energy difference is a factor of ten to the fifth. And this is mind-blowing." > — Loss summarizing the device-level energy advantage of a spin-qubit Toffoli gate over its CMOS equivalent, after accounting for refrigeration overhead. Insight: Loss argues that the Quantum Zeno effect — the tendency of frequent measurement to freeze a quantum state — can be used deliberately to stabilize classical spin states against relaxation, turning a well-known quantum-computing obstacle into a memory-stabilization tool for classical logic. Insight: The proposal requires superposition only inside a gate operation, not between operations. This means the error-correction burden is classical (majority voting) rather than quantum (surface codes or similar), which is a qualit...

Román Orús is one of the rare physicists who built a foundational mathematical tool — tensor networks — and then watched it become the engine of a unicorn. His 2013 introduction to tensor networks has been cited over 2,000 times; his company, Multiverse Computing, just announced a $570 million Series C at a $1.7 billion pre-money valuation. That arc — from condensed matter theory to Europe's largest quantum software company — is worth understanding on its own terms. But what makes this conversation particularly timely is a May 2026 paper Orús co-authored demonstrating that individual layers of Meta's Llama 3.1 8B language model can be encoded as quantum circuits and executed on IBM's 156-qubit Quantum System Two while the model generates text. It's a proof of concept, not a product — but it's a real result, and Orús is honest about what it does and doesn't prove.This episode is for listeners who want a technically grounded, hype-free account of the quantum-AI intersection: what tensor networks actually are, why they keep getting rediscovered across different fields, where classical simulation of quantum systems genuinely competes with quantum hardware, and what it looks like to build a company at the boundary between those two worlds.Sponsor MessageThe Capital of Quantum is a people story. Built on a top-five quantum PhD program  and 35-plus years of quantum research leadership. It's the billion-dollar initiative behind Discovery Center, launching this month with Microsoft, IQM, and Quantum Motion inside. That's why IonQ was born and is headquartered here, and why global companies keep choosing a spot minutes from Washington, D.C. This is where quantum is transforming the world. Come see it at the Quantum World Congress, September 23rd through 25th, College Park, Maryland. CapitalOfQuantum.com.What We Get IntoWhat tensor networks actually are — Orús explains the core idea without equations: tensors as the "DNA" of a quantum state, and how a network of them lets you see and quantify the internal correlations (entanglement) that matter versus the ones you can safely ignore.Why the same math keeps appearing in different fields — from condensed matter simulation to quantum computing simulation to machine learning, and why Orús sees that recurrence as a sign of something deep rather than a coincidence.How ChatGPT changed Multiverse's trajectory — the company was already applying tensor networks to machine learning before 2022; the emergence of large language models gave them a problem where the fit was obvious and the market was enormous.What "90–95% compression with minimal accuracy loss" actually means — Orús explains the overparameterization problem in current AI models and why he believes tensor networks address a genuine structural inefficiency, not just a tuning opportunity.The IBM kicked Ising model episode — Orús describes how his team rapidly produced a classical tensor network simulation of an experiment IBM had presented as evidence of quantum utility, and what that kind of competition between classical and quantum methods actually does for the field.The Cayley Unitary Adapter experiment — how Multiverse sliced individual layers out of Llama 3.1 8B, encoded them as quantum circuits, ran them on a 156-qubit IBM processor, and achieved a 1.4% perplexity improvement — and why Orús argues the improvement-per-parameter ratio is the number that matters, not the headline percentage.Why edge deployment is the real commercial driver — drones, satellites, vehicles, and industrial devices that cannot rely on cloud connectivity are the market pulling Multiverse toward smaller, more efficient models, not just benchmark competition with frontier labs.How Orús thinks about Multiverse's identity — he calls it a "quantum AI company," not a quantum company or an AI company, and explains what that distinction means for how they allocate research effort and where they expect to be when fault-tolerant quantum hardware matures.What he'd tell a PhD student today — a genuinely honest answer about the trade-offs between academic research and deep-tech industry, from someone who has lived both simultaneously.Resources & LinksGuest & CompanyRomán Orús — Personal Site — Lists talks, reviews, affiliations, and awards including the 2024 Physics, Innovation and Technology Prize from the Royal Spanish Society of Physics.Multiverse Computing — Official Website — Home page for CompactifAI, Singularity, and Multiverse's full product portfolio.Papers & Articles Discussed in This EpisodearXiv 2605.05914 — "Quantum-enhanced Large Language Models on Quantum Hardware via Cayley Unitary Adapters" (May 2026) — The paper at the center of the episode: Llama 3.1 8B layers running on IBM's 156-qubit Quantum System Two. Start here if you want the technical details behind the quantum-in-an-LLM result.Multiverse Computing — "Talking to a Quantum Computer" (May 2026) — The accessible blog-post version of the Cayley Unitary Adapter experiment; a good entry point before tackling the arXiv paper.arXiv 2401.14109 — CompactifAI: Extreme Compression of Large Language Models Using Quantum-Inspired Tensor Networks — The foundational peer-reviewed paper introducing Multiverse's tensor network compression method.arXiv 2509.06653 — "Classical Neural Networks on Quantum Devices via Tensor Network Disentanglers" (Sep 2025, rev. Apr 2026) — Orús, Singh, and Aizpurua on hybrid classical-quantum execution for bottleneck neural network layers; the research underpinning the longer-term hybrid architecture vision.Models & ProductsHyperNova 60B on Hugging Face — Open-source 50%-compressed model derived from GPT-OSS-120B; context for Multiverse's open model strategy.Pulsar 16B Launch with NVIDIA (June 2026) — Announcement of Multiverse's open reasoning model, scoring 87.22 on AIME 2025 at 16B parameters.Funding & Company ContextSeries C Announcement — GlobeNewswire (July 2026) — $570M / €500M round at $1.7B pre-money valuation; covers investor lineup and deployment verticals.Key Quotes & Insights> "We are using atomic bombs to kill a mosquito." Orús on the overparameterization of current large language models — and why he believes the transformer-attention paradigm, however successful, ...