Skip to content
PodSubs
ShowsPeopleGuestsJobsNewsEvents
More
BlogPodCred
Menu
ShowsPeopleGuestsJobsNewsEventsBlogPodCredNetworksSign in
Join nowSign in
  • Shows
  • People
  • Jobs
  • News
  • More
    • Guests
    • Events
    • Blog
    • PodCred
    • Networks
    Join nowSign in

Quickly Quantum

Quickly Quantum artwork

IonQ vs. Rigetti: Trapped Ions or Speed to Useful Quantum? — September 16, 2026

Sep 16, 202615 minAudio

IonQ and Rigetti Computing get put head-to-head on accuracy versus speed, Beijing rolls out a quantum-industry matchmaking program, Silicon Quantum Computing hires a semiconductor-industry CFO, Mitsubishi Electric backs quantum startup OptQC, and Multiverse Computing debuts an AI model built partly on quantum-generated data. IonQ vs. Rigetti Computing: Which Pure-Play Quantum Stock Actually Has the Technology Edge? - The Motley Fool Beijing Launches Program to Connect Quantum Companies With Industrial Customers - The Quantum Insider SQC Appoints John Hollister as Chief Financial Officer - The Quantum Insider Mitsubishi Electric invests in quantum computing startup OptQC - Evertiq CompactifAI's Quasar 1.1 438B is first AI rebuilt with quantum inputs - Quantum Zeitgeist Source links @MultiverseCompu on X fool.com Source links Source links @IonQ_Inc on X @IonQ_Inc on X Quickly Quantum is an AI-voiced podcast, built and run by a real person. Nothing in this episode is financial advice. 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.

Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

Transcript

From the show.

Which architecture actually gets you to useful quantum computing first — the one built from ultra-precise, individually trapped atoms, or the one built for raw speed? Today on Quickly Quantum: a fresh technical head-to-head pits IonQ against Rigetti Computing, and the verdict isn't nearly as settled as either company's marketing wants you to believe. Before that, in the headlines: Beijing launches a government matchmaking program to get quantum companies in front of real industrial customers, Silicon Quantum Computing brings in a semiconductor-industry veteran as its new CFO, Mitsubishi Electric takes a stake in a Japanese quantum startup, and a Spanish AI company says it's built the first model trained partly on data generated by an actual quantum computer. Welcome back to Quickly Quantum, your daily brief on the quantum frontier. It's Wednesday, September 16, 2026. Let's get into it.

Beijing wants quantum computing out of the lab and onto factory floors, and it just built the bureaucratic pipeline to make that happen. The city government launched what it's calling the Scenario Handshake Plan, a matchmaking program pairing quantum companies and university research teams with actual industrial customers. According to The Quantum Insider, officials presented six prospective applications at the unveiling: macroeconomic forecasting and tax-compliance modeling, underwater sensing and imaging, quantum-secured communications for vehicles, electrical substations, crude-oil transportation scheduling, and materials simulation, unveiled at Beijing's 2026 Industrial Future Conference, held September ninth and tenth in the city's Yizhuang development district. That district already claims more than sixty companies working across superconducting, trapped-ion, neutral-atom, and photonic quantum systems — different physics, same neighborhood — and more than thirty of them have signed onto a related industrial group called Quantum Constellation. Feng Dagang, CEO of the conference organizer 36Kr, put the stakes bluntly, saying quantum computing's return cycle, quote, 'is measured in decades, with no quick money to be gained.' The event also unveiled a new Beijing Quantum-Classical Integration Innovation Center, controlled by China Telecom Quantum Group. So is this the moment quantum research finally meets a real customer? Not yet — none of these use cases have shown an actual quantum advantage over classical computers, these are proposed demonstrations, not finished deployments. Now, this is single-source reporting out of Beijing, so we haven't independently confirmed the scope of it ourselves.

Silicon Quantum Computing just hired a chip-industry money man. SQC named John Hollister Chief Financial Officer, effective August seventeenth, after more than two decades in semiconductor finance — he was CFO at GlobalFoundries, one of the world's largest chip foundries, and before that spent two decades at Silicon Labs, including eleven years as the company's CFO there. SQC's CEO, Michelle Simmons, said Hollister brings, quote, 'financial rigor and expertise that is critical to advancing our commercial roadmap.' The hire follows SQC bringing on a Chief Legal Officer back in June, and it lands as SQC — one of eleven companies selected for Stage B of DARPA's Quantum Benchmarking Initiative — keeps pushing its atomic-precision manufacturing approach toward commercial-scale machines. Now, when a quantum startup goes and hires someone who spent two decades in the books at an actual chip foundry, that tells you something about which phase of the business they think is coming next.

Mitsubishi Electric is putting money behind a quantum startup called OptQC, taking a stake in the company. The financial terms haven't been disclosed, and the available reporting doesn't spell out exactly what Mitsubishi Electric plans to do with the technology once it's in hand. Now, you don't need the dollar figure to read the signal: when an industrial conglomerate the size of Mitsubishi Electric decides quantum computing is worth writing a check for, that's a data point on where big manufacturing thinks this technology eventually pays off, even before anyone can say precisely how or when.

Multiverse Computing says it's built the first AI model trained in part on actual quantum-computer output — not quantum-inspired math tricks, but real circuits run on IBM's 156-qubit Heron processor in Donostia-San Sebastian, Spain. The resulting model, Quasar 1.1 438B, is a rebuilt version of the open-weights GLM-5.2 model, and the company says a portion of Quasar 1.1's training relied on a so-called 'healing set' generated on that quantum hardware. Separately, in related work described alongside the release, Multiverse says quantum-generated data shaped roughly one-sixth of the layers in a Qwen3-30B-A3B model — a different model from Quasar 1.1 itself, but a further illustration of how deep this integration goes. The company's own language was direct: quote, 'Quantum computing is not a label on this release, it is part of how the model was built.' Now, the numbers on Quasar 1.1 are genuinely striking — output length dropped 37.6%, from an average of 3,322.9 tokens down to 2,074.3, while the model gained 6.2% on the HLE benchmark, 4.3% on GPQA, and 4.6% on instruction-following. It also slashed refusals on politically sensitive prompts, from 71.18% under the original GLM-5.2 down to 41%, while holding a 93% refusal rate on genuinely harmful prompts. So how much of that improvement actually traces back to the quantum-generated slice of the data, versus just having more training data generated any way at all? Multiverse doesn't say — and nobody's independently benchmarked that yet.

Our main story today has a thesis: the fight between IonQ and Rigetti Computing isn't really about which stock is cheaper right now — it's about which quantum architecture actually gets to something useful first. Quick reset if you're new to this: IonQ builds its quantum computers out of trapped ions — individual charged atoms held in place by electromagnetic fields and controlled with lasers. Rigetti builds superconducting qubits — tiny circuits cooled to near absolute zero that behave like artificial atoms. Different physics, different tradeoffs, and for a while now they've been the two most-watched publicly traded, pure-play quantum companies — meaning, unlike a giant such as IBM or Alphabet, quantum computing isn't a side project buried inside a bigger balance sheet for either of them. It IS the balance sheet. So every technical head-to-head like this one gets read by investors as a proxy battle for which stock is the safer bet, even when the underlying science hasn't actually settled that question. We've talked about IonQ on this show before — most recently when the company's own estimates about a future twenty-thousand-qubit machine cracking Bitcoin's encryption made headlines earlier this month. Today's story is a different angle on the same company: not what it might someday do to cryptography, but whether its core hardware approach is actually ahead of its biggest rival right now. According to a fresh technical comparison from The Motley Fool, IonQ still holds the edge on two fronts: accuracy, and qubit connectivity — that's the ability of qubits to interact directly with each other rather than needing extra operations to bridge them, which matters because more direct connections generally mean fewer errors creeping into a calculation. Context matters here, and it's the kind of detail a scorecard glosses over. IonQ's next-generation system, called Superion 256, is still a prototype — it hasn't shipped as a commercial product. So one side of this comparison is measuring a roadmap promise against the other side's shipped, working hardware, which is worth sitting with before you let anyone declare a winner. Now, before going further, the caveat that matters most: this comparison comes from one outlet's read of numbers that IonQ and Rigetti each disclosed about their own machines. It isn't an independent, apples-to-apples benchmark run by a neutral lab — it's directional, built on company-supplied figures, and neither company had a reason to make its rival look good. That leaves two real open questions. First: if Rigetti keeps closing the accuracy gap while holding onto its speed advantage, does the entire competitive picture flip? Second: how much of IonQ's connectivity edge survives once Superion 256 actually leaves the prototype stage and has to perform in the real world instead of on a spec sheet?

So how does the wider field see IonQ's technical standing right now? That's IonQ talking about IonQ, not an independent judge of the hardware rivalry — but it's the company's own framing of its research recognition, and it's worth keeping that separate from either side's benchmark marketing. Here's my read on the whole comparison. Speed versus accuracy is a real, live tradeoff in quantum hardware — not a marketing invention. It shows up in error rates, it shows up in how many operations survive before noise wrecks the answer, and it's exactly the tradeoff that will eventually decide which architecture gets to something practically useful first. So the rivalry itself isn't manufactured. What I'm not ready to do is crown a winner off this piece alone. The numbers behind 'IonQ is more accurate' and 'Rigetti is faster' both trace back to each company's own disclosures about its own machine, filtered through one financial outlet's analysis — not a controlled, third-party benchmark where both systems run the identical problem side by side. Every time I've seen a hardware comparison built entirely from vendor-supplied numbers, the vendor with more to prove tends to have picked the metric that flatters it. And there's a structural mismatch in this comparison that the story surfaces without fully resolving. IonQ's Superion 256 is still a prototype. So which system actually wins the accuracy-versus-speed argument once Superion ships and has to hold up outside a lab instead of on paper? Nobody knows yet — probably not even IonQ. The mirror-image question matters just as much: if Rigetti manages to close its accuracy gap while keeping its speed edge, this whole pure-play rivalry narrative — the one shaping how both stocks get talked about — flips with it. Worth spelling out why connectivity actually matters, for newcomer and expert alike: on a chip where qubits can only talk to their nearest neighbors, running an algorithm that needs distant qubits to interact means adding extra swap operations — and every extra operation is another chance for error to creep in. That's the practical case for why more direct connections, IonQ's advantage here, could translate into cleaner answers on genuinely hard problems, if it holds up at scale. If you're an investor weighing these two stocks off a headline like this one, the honest takeaway is that you're choosing between a company with a shipped, if less accurate, machine, and a company betting its edge holds once its next-generation hardware actually leaves the lab. Now, why should any of this matter to someone who doesn't own either stock? Because whichever architecture actually solves the speed-versus-accuracy tradeoff first is the one likely to land the first genuinely useful commercial contract — the kind of contract that decides whose approach gets replicated at other companies and whose gets left behind as a footnote. One more thing before we move on: there's a bigger financial story bubbling under the quantum foundry world right now — whether IBM's roughly one-billion-dollar CHIPS Act award for its own quantum chip foundry has actually been finalized. We're not calling that settled today, because the reporting we could verify doesn't clearly confirm it, but we'll dig into it as soon as it firms up. Time for the Hype Check. I'm putting this one at a 4 out of 10 on substance. The underlying physics tension — trapped-ion precision against superconducting speed — is real, and it's worth following closely. But this specific piece is one outlet's directional read of numbers both companies supplied about themselves, dressed up as a head-to-head verdict. Call it a scorecard, not a benchmark.

So here's the listener lens for today: next time you see a company claim it's 'ahead' in a two-way quantum race, ask which axis it picked to win on — accuracy, speed, or raw qubit count — before you buy the whole scoreboard. If you want tomorrow's episode the moment it drops, including whatever we can confirm on that IBM foundry story, follow Quickly Quantum wherever you're listening. 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!

I also host Space Stakes: the business of the new space race, every day. What actually flew, what the contract is really worth, and who has customers. Find it wherever you get your podcasts.

Open the transcript file

Hosts and Guests

Nobody is credited on this episode yet.

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. Willow Tests Measurement Contextuality at 72- and 105-Qubit Scale — November 20, 2026

    Nov 20, 202618 minAudio

    Explainer: harvest now, decrypt later - Wikipedia memQ and Infleqtion Sign MoU to Advance Networked Quantum Computing - Infleqtion SEALSQ Signs LOI to Acquire QuintessenceLabs - The Quantum Insider Chapters 00:00...

    Credit the people on this episode

  2. Finance Gets an Immediate Quantum Warning — November 19, 2026

    Nov 19, 202617 minAudio

    A global securities regulator has urged the financial sector to act on quantum risk, with blockchain systems singled out as potentially exposed. We also cover hardware roadmaps, location verification, quantum...

    Credit the people on this episode

  3. IQM’s 150-Qubit Start Comes With an Upgrade Promise — November 18, 2026

    Nov 18, 202617 minAudio

    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...

    Credit the people on this episode

  4. Can Quantum Chips Fit the Supercomputer? — November 17, 2026

    Nov 17, 202619 minAudio

    IBM, NVIDIA and Fermilab plan a modular quantum-classical platform, while new announcements address scheduling, error correction and quantum-safe communications. We also cover DOE testbed planning, U.S. chip-access...

    Credit the people on this episode

  5. Moth's 1,568-Value Test Takes On Quantum Data Loading — November 16, 2026

    Nov 16, 202616 minAudio

    Moth Quantum reports a larger data-loading and recovery demonstration on IBM processors, while the episode examines readout costs and the distance between device specifications and useful workloads. Other stories cover...

    Credit the people on this episode