Quickly Quantum

IQM’s Halocene programme promises continuity through hardware upgrades, while today’s other stories examine quantum-safe telecom, national-security concerns, funding and emerging computing systems. The episode separates demonstrations and company plans from longer-term claims.

Chapters

00:00 Today's headlines

00:45 IQM upgrades Halocene systems without downtime

04:15 No delivery date for the larger Halocene system

07:34 QMill quantum-powered encryption in 5G network

09:22 Israel's military edge and quantum technology

11:05 London Quantum Cluster received £12 million

12:24 PsiQuantum and AMD announce PRISM decoder

13:14 Q.ANT demonstrates photonic NPU with AMD GPU

14:43 SAXON Q introduces Quantum Watchdog

Source links
Quickly Quantum is an AI-voiced podcast, built and run by a real person. Nothing in this episode is financial advice.
Follow Quickly Quantum: subscribe on YouTube · Apple Podcasts · Spotify
More from Brian Lampert: Concrete Compute, the daily AI infrastructure briefing, and Space Stakes, the business of the new space race. Transcripts and every episode: quickly-quantum.kngoworld.chatgpt.site.

What is Quickly Quantum?

The daily quantum computing briefing you don't need a physics degree to follow. Every day, Quickly Quantum cuts through the hype to bring you the breakthroughs, funding rounds, policy moves, and research that actually matter, with plain-English analysis a smart non-physicist can follow. Hosted by Brian Lampert. AI-voiced, human-built, always skeptical of press releases. Nothing on this show is financial advice.

Also from Brian Lampert: Concrete Compute and Space Stakes.

IQM says Halocene customers can begin with a 150-qubit system and upgrade toward systems with more than 1,000 qubits. The consequential part is the promise that a new system can be installed while the current one keeps running, a practical answer to the fear that an early purchase gets overtaken. But the larger system is a future destination in the company’s announcement, not a machine demonstrated today. We’ll work through what the upgrade actually carries forward, and what the announcement leaves as a promise. Welcome back to Quickly Quantum, your daily brief on the quantum frontier. It's Wednesday, November 18, 2026. Let's get into it.

IQM is pitching Halocene as a system customers can grow with, tying the purchase of quantum hardware to an ongoing relationship with a supercomputing center’s existing software and infrastructure. The Quantum Insider reports that the company announced an upgrade programme for its error-correction product line, moving from a 150-qubit system toward a system with more than 1,000 qubits. Error correction is the set of methods used to protect quantum calculations from errors that would otherwise spoil the result; fault-tolerant computing is the longer-term goal of keeping those protected calculations reliable as a machine operates.

The announcement says customers can run hybrid workloads today alongside AI and high-performance computing systems. Hybrid here means quantum processors working alongside conventional processors, rather than replacing the CPUs and GPUs already doing AI and simulation work. IQM says customers can upgrade to each new Halocene system without downtime, and that its high-performance-computing integration service comes with every upgrade.

What does an upgrade involve in practice? IQM says it installs the new Halocene system while the current one keeps running, then removes the previous system once the new one is in service. Its integration service connects the new system to the center’s existing infrastructure, schedulers and GPU platforms. A scheduler is software that assigns computing jobs to available resources, so connecting to that environment is part of making a quantum machine usable alongside the center’s other equipment.

The company’s case rests on continuity between generations. IQM says Halocene systems share common infrastructure, error-correction methods and software across generations, allowing investment in one system to carry forward to the next. That is the substance beneath the upgrade offer: customers could preserve integrations, workflows and skills as the hardware changes, rather than treating each new machine as a fresh installation.

The Quantum Insider also reports examples IQM uses to support that case. VTT has moved through successive IQM systems since 2021, from 5 qubits to a 53-qubit system coupled with the LUMI supercomputer. The Leibniz Supercomputing Centre has integrated four IQM systems into its supercomputing infrastructure, and IQM says software, workflows and skills were carried forward in each case.

Dr. Tobias Guggemos, Quantum Technologies Team Lead at the Leibniz Supercomputing Centre, says, “Integrating quantum computers into our supercomputing environment has shown us that the real value lies in continuity. Each IQM system we have added has built on the integration, software and expertise we developed with the previous one, so our users keep working while the technology moves forward. An upgrade path designed around that continuity is what supercomputing centers need on the way to fault-tolerant quantum computing.”

That’s the clearest customer case in the announcement: continuity matters to a center that has already integrated successive systems. IQM also says Halocene is built on an open and modular error-correction stack, so customers can experiment, develop their own methods and avoid lock-in. The company says its integration work draws on its collaboration with NVIDIA, including on NVQLink for scalable quantum error correction.

There is a current deployment alongside the future-facing upgrade claim. The Quantum Insider says IQM introduced Halocene in November 2025, starting with a 150-qubit system, and that the LUMI AI Factory, coordinated by CSC – IT Center for Science, has since chosen it for LUMI-IQ, which will be connected to the upcoming LUMI-AI supercomputer. That gives the programme a concrete setting in which the hybrid approach matters, even though it does not demonstrate the future upgrade itself.

The harder question is what the larger system will mean in usable computing, because the announcement gives a qubit target but no logical-qubit specification for that milestone. A logical qubit is a more reliable unit built from physical qubits and error correction; the distinction matters because a larger count of physical qubits alone doesn’t establish how much fault-tolerant work a machine can perform.

IQM’s own description makes the upgrade programme sound operationally specific: the new system is installed while the existing one runs, then the old one is removed after the new one enters service. The company says shared infrastructure, error-correction methods and software link the generations. That’s a design for continuity, and it answers the practical concern about interruption more directly than a general promise to scale.

But “without downtime” remains a programme claim, not a result demonstrated in this announcement. The more-than-1,000-qubit system is described as the system of the future; IQM gives no delivery date for that milestone. And a qubit count is not the same thing as a measure of useful, error-corrected computing. Those are important distinctions for a buyer weighing the value of an installation today against what a later generation might offer.

IQM CEO and co-founder Dr. Jan Goetz says, Solving the physics is only half the challenge. The other half is making quantum computing work commercially, inside the infrastructure organizations already run. With Halocene now upgradable to over 1,000 qubits, our customers can start today, run hybrid workloads alongside their AI and HPC systems, and move to each new system without downtime. No one has integrated or upgraded more quantum computers than IQM, and that experience is what makes this path to fault tolerance work.

The commercial point is real: a quantum processor that fits into a center’s existing computing environment may be easier to adopt than one that stands apart from it. The company’s customer examples support the narrower claim that integrations and expertise can carry between systems. They don’t establish how capable the future Halocene will be, or how much useful error-corrected computation it will deliver.

I think this is a credible infrastructure and customer-continuity pitch, and a more grounded one than a bare promise of a bigger processor. The strongest evidence in the announcement is the described upgrade process and the examples of existing systems, not the future qubit milestone. I’m less persuaded by the leap from a shared platform to a demonstrated path to fault tolerance, because the announcement doesn’t give the logical-qubit details that would let us assess that leap.

That doesn’t make the programme empty. For a research center, the chance to keep workflows and integrations moving as hardware changes could have value before a machine reaches fault tolerance. But the company’s language about the destination should stay attached to the destination: customers can start with the 150-qubit system, while the system with more than 1,000 qubits is still ahead.

Time for the Hype Check. I’d give this a 6 out of 10 for substance: the strongest point against that score is that the larger system has no stated delivery date or logical-qubit specification, while the upgrade design and customer continuity examples give the promise something concrete to rest on.

IQM is presenting Halocene at SC26 in Chicago, November 15-20, and that presentation will show how the company explains the upgrade path, while the future system’s computing capability remains unproven.

Away from the question of upgrading machines, The Quantum Insider reports that Telia Finland, Nokia and QMill demonstrated quantum-powered encryption in a 5G network using commercial equipment. The partners say they integrated QMill’s solution into Nokia’s 5G base-station software and demonstrated it in Telia’s 5G innovation environment; this show hasn’t independently confirmed the claim.

The system combines standard ML-KEM post-quantum cryptography with QMill’s quantum-powered key exchange and user-verifiable quantum randomness. Post-quantum cryptography, or PQC, is encryption designed to withstand attacks from future quantum computers. The network’s security gateway protects traffic headed to the operator’s core network, and the demonstration used a quantum computer to establish a secure connection between that gateway and a 5G base station.

The companies describe two functions. First, the base station combines ML-KEM, a post-quantum key exchange algorithm standardized by NIST, with QMill’s new quantum-powered key exchange algorithm. Second, QMill says it uses fresh, user-verifiable random numbers from a quantum computer to seed that exchange, with the computations fetched through its QMill Cloud service.

The Quantum Insider says the proof of concept went from concept to demonstration in less than 90 days. Telia, Nokia and QMill are now exploring security-sensitive use cases for commercialization. The demonstration is on standard network equipment, and the partners say it didn’t require costly dedicated infrastructure.

For you, the important distinction is that standard PQC remains the foundation for critical networks in the partners’ description; QMill’s quantum-powered method is an added component. I think the compelling part is the integration into carrier equipment, while the next challenge is whether the partners identify a concrete use case that merits the extra component. So far, they say they’re looking for one.

The security discussion shifts from network encryption to national defense. The Jerusalem Post reports that a new Institute for National Security Studies report warns quantum technology could threaten Israel’s long-standing qualitative military advantage. A qualitative military edge means a country’s perceived technological advantage over potential adversaries; the report connects that concern to quantum computing, sensing and other technologies.

The report says Israel ranks second globally after China in quantum investment relative to GDP. It expects cumulative investment in Israel’s quantum sector to reach approximately $11 million by the end of 2026, placing Israel fifth worldwide in overall funding and seventh in quantum innovation. The article also says an October report by the Israel Innovation Authority and C4IR Israel found Israeli quantum companies raised nearly $500 million in 2025, accounting for approximately 9% of global private quantum investment and placing Israel fourth worldwide.

The INSS report points beyond computing to quantum sensing. It says quantum radar could eventually detect stealth aircraft, missiles and submarines, potentially changing some advantages held by advanced militaries. That’s a projection about a technology’s possible future use, not evidence of a deployed capability.

The report calls the race a strategic objective and says it could drive a shift toward “quantum warfare.” I think the article’s most useful distinction is between that broad warning and the concrete evidence it offers: investment rankings and a set of possible military applications. The warning matters as policy analysis; it doesn’t establish that quantum systems can currently do those things.

Meanwhile, public money and private pledges are being gathered around London’s quantum industry. Quantum Zeitgeist reports that the London Quantum Cluster, co-led by Imperial, King’s College London and UCL, received £12 million from the Mayor of London, alongside pledges of more than £50 million from 30 partners. Those are distinct figures and distinct sources of funding.

The cluster aims to grow quantum firms and jobs, and the funding announcement says it will support up to 250 jobs and 500 businesses. The project will focus on WestTech London and the Knowledge Quarter, using existing infrastructure to support testing and commercialization. The article says potential applications span healthcare, transport, finance, energy and defense.

Professor Mary Ryan, Vice-Provost (Research and Enterprise), said, “We have seen firsthand, through the WestTech London ecosystem, the power of bringing together higher education, the public sector and industry to turn research excellence into breakthroughs and businesses which create jobs, drive growth and improve people’s lives.”

The announcement also says the funding was made during London Life Sciences Week, which drew over 5,000 attendees. I’d treat the jobs and business figures as the programme’s stated reach, with the real measure arriving in whether the cluster turns its partners and infrastructure into operating companies and employment.

And on the error-correction front, PsiQuantum announced on X that its researchers and @AMD researchers introduced PRISM, a machine-learning decoder for real-time error correction in photonic quantum computers. A decoder is software that interprets error information so a system can correct faults in its computation.

PsiQuantum’s official account says the work appears in a new joint paper and is part of the companies’ recently announced strategic partnership. The post describes the research as preparation for fault-tolerant, utility-scale quantum computing. That’s the company’s characterization of the work, not a claim here that those larger systems have arrived.

The announcement adds a specific technical layer to the day’s broader error-correction story: a machine-learning decoder aimed at real-time operation in photonic machines. The paper is the concrete thing to assess; the social post is a short account of its purpose.

On a different computing path, The Quantum Insider reports that Q.ANT demonstrated its NPU Gen 2+ photonic processor alongside an AMD GPU on a language-model workload. Photonic computing uses light to carry out calculations, and the new processor adds optical discrete Fourier transforms, or DFTs, a mathematical operation used in signal and image processing, simulations and data analysis.

Q.ANT says the demonstration used optical calculations to replace a computationally intensive attention layer in a large language model. Attention is the part of a language model that helps it model relationships between tokens, or pieces of text. The company says the DFT runs natively in light, while the processor acts as a co-processor alongside conventional CPUs and GPUs.

That is a hybrid photonic-GPU demonstration, and it belongs in the wider computing conversation, though it isn’t a quantum-computing result. Q.ANT is extending its software toolkit and plans to offer commercial systems combining Gen 2 and Gen 2+ processors in the second half of 2027.

Q.ANT CEO Michael Förtsch said, “With Gen 2+, we’re opening the door to software that uses what light does naturally. That allows leaner AI models with fewer parameters and less data, and that is what will make the next wave of AI affordable.” I’d keep the claim of affordability in the company’s own voice; the demonstration establishes a workload, while the commercial systems remain planned.

A final hardware story takes up the question of what happens when a quantum computer misbehaves. The Quantum Insider reports that SAXON Q introduced Quantum Watchdog, a patented system designed to detect hardware failures and trigger protective responses. The company says implementation will be available in early 2027.

The watchdog uses its own classical processor, with a separate clock, timers and sensors, to run verified test calculations and monitor operating conditions. Instead of checking every user answer, it periodically swaps a user’s calculation for a test problem with a verified answer. It also tracks power draw, temperature and magnetic field; if results or conditions fall outside set parameters, the system can respond according to established protocols.

For a machine running two quantum processors, the watchdog tests both and compares their answers, then puts the unit into safe mode if they disagree. With three processors or more, it takes a vote and can shut down processors whose answers don’t match the majority.

SAXON Q co-founder and CEO Marius Grundmann draws a useful line: “Error correction handles errors, the watchdog catches failures.” Error correction addresses mistakes in computation; this system is aimed at hardware problems such as defects, aging or tampering. Custom configurations can devote up to half of processing time to safety checks, while fully certified versions are expected within the next few years.

That’s a distinct engineering problem from correcting faulty calculations, and potentially important for systems used beyond research settings. The company’s design is specific; independent performance evidence isn’t part of this announcement.

If this mix of quantum hardware, security and the claims in between is useful, follow today’s episode wherever you listen. This has been Quickly Quantum, an AI-voiced podcast, created and built by a real human using today's cutting-edge technology. Nothing you heard on this show is financial advice. I'm Brian Lampert, and I'll catch you all tomorrow — take care!

Watching on YouTube? Hit subscribe, and tomorrow's briefing will be waiting for you. I also host Concrete Compute: a daily briefing on the AI buildout. The datacenters, the megawatts, and who actually pays for them. Find it wherever you get your podcasts.