Quantum Advantage Battles: Qubit Supremacy and Banking Hype. Caltech goes over 3000 - The Week in Quantum Computing, September 29th 2025
Issue #253
YES YES and YES. This is the types of weeks I love to write my quantum newsletter. Quantum advantage claims happening left and right. New proper developments, and fights in the port canteen. Read on scallywags. For we are about to set sail!
This week the quantum world felt like a schoolyard brawl. IonQ came out swinging, proudly declaring it hit an “algorithmic qubit” benchmark of #AQ 64 on its new Tempo system – even crowing that it’s the only company to reach that level, “setting a new standard for quantum systems”ionq.com. In other words, IonQ flexed that its qubits are the biggest and baddest (with claims of broad quantum advantage on the horizon). Quantinuum quickly clapped back, essentially yelling “cheater!” across the yard. In a scathing blog post, they argued IonQ’s beloved metric is “deeply flawed, hiding computer performance behind a plurality voting trick and gate compilations that are not widely useful”quantinuum.com. It’s like one kid bragging about a video game high score after using cheat codes, while the other insists on playing by official rules (Quantinuum prefers industry-standard Quantum Volume benchmarks) and having the real high score. The result? An entertaining spat over who’s got the better qubits, with each sibling convinced the other’s trophy is made of fool’s gold.
Meanwhile, HSBC (yes, the bank) decided to join the fray with some chest-thumping of its own. Together with IBM, HSBC trumpeted a “ground-breaking world-first” example of quantum computing in finance, claiming a trial that gave a 34% boost in bond trading predictions – “the world’s first-known empirical evidence” that today’s quantum hardware can solve a real business problem and confer a “competitive edge”hsbc.comhsbc.com. If that sounds like every hype buzzword thrown into a blender, well, cue the skeptics. Enter Scott Aaronson – one of those salty old sailors of science not easily swayed by marketing sirens. Aaronson promptly dubbed HSBC’s claim a “qombie” (as in zombie quantum advantage): an undead claim that refuses to die. He noted the paper raised “more red flags here than in a [military] parade,” deeming the entire study “fatally flawed…optimized for generating headlines rather than knowledge”scottaaronson.blog. The supposed advantage, he pointed out, only appeared because the quantum processor’s noise made the data look better – when a noiseless simulator was used, the advantage vanished, “nothing really to do with quantum mechanics” at all. In his words, “qombies roam the earth: undead narratives of ‘quantum advantage for important business problems’ detached from any serious underlying truth-claim”scottaaronson.blog. Ouch. HSBC’s boast got a harsh reality check, basically being told that their quantum crown jewel is a cubic zirconia at best.
All told, it’s been a wild week of quantum mud-slinging – and a reminder that this industry hasn’t outgrown its playground antics. Bold claims of “quantum advantage” keep popping up like whack-a-mole, only to be whacked back down by even bolder rebuttals. It’s nothing we haven’t seen before: “loud declarations of various types of quantum advantage aren’t new”, and past announcements (Google in 2019, IBM in 2023, D-Wave, you name it) “were ultimately refuted by outside researchers who used clever classical techniques” to nullify the supposed breakthroughsscientificamerican.com. Even when a claim is legitimate in a technical sense, it often “has no immediate commercial application” – remaining a proof-of-concept that “has yet to solve real-world problems”dig.watch. In other words, lots of chest-beating and press releases, but not a lot of actual booty to show for it. The hype-happy pirates of quantum computing might all be dueling for the spotlight, but so far the treasure at stake looks more like fool’s gold than the real thing.
What else has happened? (EVERYTHING!)
Researchers at Caltech have engineered a neutral atom device with 6,100 ultracold caesium atom qubits, surpassing the previous 1,180-qubit record from Atom Computing (similar tech as QuEra as well). They demonstrated stable qubit arrays and tested 2D laser-based qubit transport, though full computation has yet to be shown. This sets a new benchmark for quantum system scale. William Kretschmer et al. at Quantinuum demonstrated unconditional quantum advantage using Quantinuum’s H1-1 trapped-ion processor, achieving 99.941(7)% median two-qubit partial-entangler fidelity. They constructed a task with provable quantum-classical information separation, achieving “quantum information supremacy” without relying on unproven complexity assumptions, directly evidencing advanced entangled state manipulation Quantum Computing Inc closed an oversubscribed $500 million private placement of common stocl. Krown Technologies & Quantum eMotion completed development of Qastle, the first quantum-safe hot wallet integrating Quantum eMotion’s ultrafast Quantum Random Number Generator and post-quantum encryption for blockchain applications. IonQ demonstrated quantum frequency conversion for trapped ion photonic qubits to the 1550 nm telecom band with negligible state change, enabling quantum information transmission over existing fiber infrastructure. IonQ claims this positions them for leadership in quantum networking and distributed computing. Qilimanjaro & Qblox announced a strategic partnership to accelerate hybrid digital-analog quantum computing. The collaboration combines Qblox’s quantum control hardware with Qilimanjaro’s processors and software, targeting scalable and versatile hybrid quantum architectures.
HSBC & IBM delivered the first quantum-enabled algorithmic trading application using live production-scale data, with a claim of improving prediction accuracy by up to 34% in European corporate bond trading versus classical models. This establishes empirical evidence for quantum advantage in financial services. Then at the same time IonQ claims 64 Algorithmiq Qubits while Quantinumm shouts “what the hell is AQ?”
The Week in Quantum Computing
Paper: Demonstrating an unconditional separation between quantum and classical information resources
A team led by William Kretschmer has demonstrated an unconditional quantum advantage using Quantinuum’s H1-1 trapped-ion quantum processor, achieving a median two-qubit partial-entangler fidelity of 99.941(7)%. The group constructed a computational task where the most space-efficient classical algorithm provably requires between 62 and 382 bits of memory, while the quantum solution used only 12 qubits. Unlike prior forms of quantum or computational supremacy, this result—termed “quantum information supremacy”—does not depend on unproven complexity conjectures and provides direct evidence that current quantum systems can generate and manipulate highly complex entangled states. This establishes a new benchmark for quantum computing, marked by an unconditional separation in information resources between quantum and classical methods.
(Note from the pirate: advantage != useful)
Quantum Computing Inc. Announces $500 Million Oversubscribed Private Placement of Common Stock Priced at the Market Under Nasdaq Rules
Quantum Computing Inc. completed a $500 million oversubscribed private placement of common stock, priced at market under Nasdaq regulations. This significant capital raise demonstrates strong investor demand and enhances the company’s resources for advancing its business objectives. The offering’s oversubscription indicates enthusiasm for the company’s direction and market presence. The transaction involved the sale of common stock directly to private investors, raising $500 million in total proceeds. This funding strengthens Quantum Computing Inc.’s balance sheet, positioning it for future technology development and market expansion. The key achievement is the successful closing of an oversubscribed $500 million private placement, confirming robust investor support for Quantum Computing Inc.’s strategy.
Krown Technologies and Quantum eMotion Complete Development of the World’s First Quantum-Safe Hot Wallet
Krown Technologies and Quantum eMotion have completed development of Qastle, the world’s first quantum-safe hot wallet. This wallet integrates Quantum eMotion’s patented ultrafast Quantum Random Number Generator (QRNG), producing true entropy at gigabit speeds, and overlays post-quantum encryption protocols. This technical advance ensures cryptographic keys are resistant to both classical and quantum attacks, directly addressing urgent blockchain security threats. According to the announcement, the global crypto wallet market was valued at $3.2 billion in 2024 and projected to reach $33.6 billion by 2033, with hot wallets comprising over half. Krown CEO James Stephens and Quantum eMotion CEO Dr. Francis Bellido emphasized the practical impact for both retail and institutional users. Key achievement: First completion of a quantum-safe hot wallet for real-time blockchain transactions.
Because you need tools to solve problems you didn’t have to start with. But better with a “Q” and blockchain around its neck.
Quantum Computing Companies in 2025 (76 Major Players)
A comprehensive 2025 survey identifies 76 major companies advancing quantum computing globally. The list spans hardware, software, and service providers, including industry leaders such as International Business Machines Corporation, Google Quantum AI, Microsoft, Rigetti Computing, D-Wave Quantum Inc., IonQ Inc., and Alibaba Group Holding Limited. These firms drive development of quantum processors, cloud platforms, and quantum-inspired solutions for sectors like finance, healthcare, and transportation. The presence of such a wide array of players underscores heightened investments and competition, accelerating innovation cycles.
IonQ Demonstrates Quantum Frequency Conversion to Telecom Wavelengths
IonQ has demonstrated quantum frequency conversion of trapped ion photonic qubits to telecom wavelengths, a key advance for quantum networking. This technical milestone enables transmission of quantum information over existing fiber-optic infrastructure, addressing a major barrier for scalable quantum internet. According to IonQ, photonic qubits from its trapped ion systems were converted to the standard 1550 nm telecom band “with negligible change to quantum state.” CEO Peter Chapman noted this paves the way for networking quantum computers over metropolitan and long-haul distances. IonQ’s work shows direct compatibility between their trapped ion technology and telecommunications infrastructure. The achievement positions IonQ for leadership in future quantum networks and distributed quantum computing, marking a significant step for quantum internet development.
Qilimanjaro and Qblox Announce Strategic Collaboration to Advance Hybrid Digital
Qilimanjaro and Qblox have announced a strategic collaboration to accelerate the development of hybrid digital-analog quantum computing systems. The two companies will integrate Qblox’s modular quantum control hardware with Qilimanjaro’s Qibo software and quantum processors. This partnership aims to facilitate seamless operation of hybrid architectures, enabling scalability and enhanced algorithm implementation. Key companies involved are Qilimanjaro Quantum Tech, known for their superconducting and spin qubit processors, and Qblox, a provider of quantum control equipment. The integrated solution is designed to address challenges in hybrid digital-analog quantum computing and support research and industrial deployment.
Device with 6100 qubits is a step towards largest quantum computer yet
Researchers at the California Institute of Technology have assembled a device with 6,100 ultracold caesium atom qubits, surpassing the previous record holder, Atom Computing’s 1,180-qubit device. The team engineered their neutral atom array for stability and extended coherence using precision laser controls, though computation has not yet been demonstrated. Laser-based transport of qubits within the 2D grid, a critical step for computation, was successfully tested. Mark Saffman at the University of Wisconsin-Madison highlights this as proof of scalability for neutral atom quantum systems, though further experimental validation is needed. Team member Kon Leung states the group aims to scale to one million qubits within a decade. This establishes a new technical benchmark for large, highly organized quantum architectures.
Caltech Team Sets Record with 6,100-Qubit Array
A Caltech team led by Professor Manuel Endres has set a new benchmark by assembling the largest neutral-atom qubit array to date: 6,100 cesium atom qubits trapped using optical tweezers arranged in a laser-generated grid. Previous neutral-atom arrays only reached the hundreds. Despite scaling up, the team maintained superposition states for approximately 13 seconds—nearly 10 times longer than previous efforts—and achieved 99.98 percent manipulation accuracy. Graduate students Hannah Manetsch, Gyohei Nomura, and Elie Bataille led the experimental work, demonstrating precise movement of individual atoms across hundreds of micrometers while preserving coherence. This achievement demonstrates both scalability and robustness, positioning neutral atoms as a strong route toward large, error-corrected quantum computers.
HSBC demonstrates world’s first-known quantum-enabled algorithmic trading with IBM
HSBC, in collaboration with IBM, demonstrated the world’s first-known quantum-enabled algorithmic trading application using live production-scale trading data. The trial yielded up to a 34% improvement in predicting the probability of winning customer inquiries in the European corporate bond market versus traditional classical techniques. (Note from the pirate: WHAT TRADITIONAL TECHNIQUES?) HSBC and IBM performed this demonstration by integrating quantum and classical computing resources to optimize quotes in over-the-counter bond trading scenarios. Philip Intallura, HSBC Group Head of Quantum Technologies, stated the experiment provides tangible evidence that “today’s quantum computers could solve a real-world business problem at scale and offer a competitive edge.”
And the debunk:
Scott Aaronson - HSBC unleashes yet another qombie: a zombie claim of quantum advantage that isn’t
HSBC and IBM released a paper claiming a 34% gain in financial trade fill prediction using data transformed by a quantum algorithm on IBM Heron processors, compared to classical models and classical simulations. However, Scott Aaronson highlights that the reported advantage only appears when using noisy quantum hardware; the benefit vanishes under noiseless (ideal) quantum simulation. The result: quantum noise, not quantum computational speedup, appears responsible for the performance change.
Highly recommended article, and the comments on Aaronson’s post where there is a very interesting debate on the HSBC piece.
Sharing learnings from the International Conference on PQC and AI
The International Conference on Post-Quantum Cryptography and Artificial Intelligence, hosted by Professor Bill Buchanan at Edinburgh Napier University, gathered leaders from Bank of Israel, Santander, and others to address post-quantum cryptography migration. Key insight: cryptographic migration requires 5–10 years, with critical “high-risk” algorithm deprecation milestones by 2030–2035. The conference highlighted live PQC deployments (such as Santander’s TLS 1.3/Openbank) and techniques for cryptographic asset discovery, like Crypto BOMs and network sensors. Hardware security modules need recertification for PQC, which may disrupt FIPS compliance. Migration strategies must prioritize use-case inventories, pilot hybrid crypto, and iterative assurance. Quantum-readiness should be measured by coverage, velocity, and risk KPIs, supporting migration dashboards and targeted compliance by 2026, 2030, and 2035.
FS-ISAC Urges Global Coordination for Migration to Post-Quantum Cryptography in Financial Services
FS-ISAC has released “The Timeline for Post Quantum Cryptographic Migration,” urging global financial institutions to coordinate their transition to quantum-resilient cryptography. Developed with experts from the Quantum Safe Financial Forum and CFDIR Quantum-Readiness Working Group, the white paper outlines key milestones, a phased transition framework, and stresses the need for synchronized planning among payment networks, cloud providers, and vendors. Mike Silverman, FS-ISAC’s Chief Strategy and Innovation Officer, emphasized the necessity of proactive, coordinated strategies. The paper supports actionable guidance for mitigating migration risks and maintaining regulatory compliance. FS-ISAC members represent $100 trillion in assets across 75 countries. This marks a sector-wide push for clear, aligned action plans to safeguard global financial infrastructure against quantum threats.
Singapore Start-up’s Quantum Cryptography Secures Data
Squareroot8 Technologies, a National University of Singapore spin-off, has launched a chip-based quantum random number generator (QRNG) designed to secure critical information systems against quantum-enabled attacks. Backed by $3 million from the National Research Foundation and US$3 million in private seed funding, the start-up’s technology leverages quantum mechanics to produce encryption keys with maximal randomness. Trials are now in discussion with firms in healthcare, defence, and finance. The company, employing about 10 engineers, aims to double staff in two years and is partnering with Danish cryptography firm Partisia to serve clients handling sensitive data for artificial intelligence. Since 2002, Singapore has directed $700 million towards quantum research, underscoring robust support for commercialization of quantum-safe technologies.
IonQ Achieves Record Breaking Quantum Performance Milestone of #AQ 64
IonQ announced it achieved an algorithmic qubit (AQ) benchmark of #AQ 64, setting a new record in quantum computing performance. This metric was independently validated by the Quantum Economic Development Consortium using IonQ hardware and the industry’s only published AQ algorithmic benchmark. Previous IonQ systems achieved #AQ 29 and #AQ 35. The AQ 64 milestone signifies the system can execute 385 qubit operations in a circuit of depth 2048, maintaining accuracy throughout. IonQ’s achievement demonstrates substantial advancement and scalability in hardware capability, progressing rapidly towards practical utility. According to the company, this result shows IonQ’s system performance has doubled approximately annually.
Note of the pirate: Only IonQ uses and accepts AQs as a valid benchmark. See below:
Debunking algorithmic qubits
Quantinuum released a detailed analysis on March 1, 2024, criticizing the algorithmic qubits benchmark as a flawed measure of quantum computer performance. Their simulations show that algorithmic qubits results can be significantly distorted by gate compilation and plurality voting—techniques that artificially boost scores and do not reflect real-world performance. Quantinuum’s assessment demonstrates that the algorithmic qubits benchmark consistently overestimates capability compared to quantum volume, a more robust and established metric. The company emphasizes that algorithmic qubits are not equivalent to logical qubits required for fault-tolerant computing. By highlighting these vulnerabilities and using data from their H2-1 system, Quantinuum reinforces that algorithmic qubits present misleading comparisons and should not replace rigorous benchmarks like quantum volume.
Proven quantum advantage: Researchers cut the time for a learning task from 20 million years to 15 minutes
A collaborative research team has demonstrated a quantum computing milestone by solving a complex learning task in 15 minutes—a process estimated to require 20 million years on a classical supercomputer. The study, led by Helmholtz Munich and published in Nature, leveraged IBM’s 127-qubit Eagle quantum processor to run a quantum kernel evaluation for a machine learning classification problem. The result showcases exponential speed-up and practical viability for quantum-enhanced machine learning, setting a benchmark for real-world quantum advantage. According to the published data, this experiment involved executing over 600,000 quantum circuits within a controlled environment.
Cisco rolls out software aimed at connecting a quantum computing cloud
Cisco has unveiled new software designed to connect disparate quantum computing clouds, according to Reuters. The rollout targets the integration of quantum computers across different cloud services, supporting interoperability and expansion of quantum networks. The announcement reflects Cisco Systems Inc’s strategic push into quantum networking infrastructure as global competition intensifies. Key figures and financial details were not specified in the article. By enabling seamless cloud connectivity, Cisco aims to bolster the development and deployment of quantum computing capabilities within enterprise and research environments. The key achievement is the introduction of software facilitating cross-cloud quantum computing integration.
Clearing significant hurdle to quantum computing
Harvard physicists, in collaboration with Massachusetts Institute of Technology and QuEra Computing, have demonstrated the first continuously operating quantum machine exceeding 3,000 quantum bits. Published in Nature, the system ran for over two hours—overcoming the longstanding challenge of “atom loss” by using optical lattices and tweezers to reload up to 300,000 atoms per second without interrupting computation. The device sustained quantum operations while cycling through more than 50 million atoms. Led by professors Mikhail Lukin, Markus Greiner, and Vladan Vuletic, the team’s innovation allows indefinite operation and dynamic reconfiguration. Federal support included the U.S. Department of Energy and National Science Foundation.
New Quantum Fault-Tolerance Approach Could Accelerate Post-Quantum Cryptography (PQC) Timeline
Researchers from QuEra, Harvard University, and Yale University introduced the Transversal Algorithmic Fault Tolerance (AFT) framework, published in Nature (“Low‑Overhead Transversal Fault Tolerance for Universal Quantum Computation”, Zhou et al., 2025), which reduces quantum error-correction runtime overhead by an order of magnitude. AFT enables logical operations with only a constant number of syndrome extraction rounds, rather than rounds proportional to code distance, achieving 10–100× speedups in large-scale quantum algorithms. This framework particularly benefits neutral-atom quantum computers, turning their slower gate speeds into a competitive advantage. A companion study shows Shor’s algorithm could factor a 2048-bit RSA key in 5.6 days with 19 million qubits—50× faster than previous estimates. The result directly tightens post-quantum cryptography timelines and compels risk planners to revise quantum readiness strategies.
In which I lose faith in quantum computing
Quantum Computing: Overvalued And Not Much To Show For
Quantum computing companies have seen skyrocketing valuations despite limited technical milestones, according to Seeking Alpha. The analysis highlights that market leaders like IonQ and Rigetti Computing report minimal revenue—IonQ posted just $11.1 million in 2022, while Rigetti generated less than $13.1 million. These figures pale in comparison to their nearly $900 million and $100 million respective market capitalizations. The article stresses a lack of commercially viable breakthroughs, citing current machines’ limited capabilities. Seeking Alpha urges caution, observing that investor enthusiasm is outpacing near-term business or technological returns. Key implication: Quantum computing’s financial hype is not yet matched by real-world progress or revenue.