Latest / Quickly Quantum / July 25, 2026 — Weekly Recap: IBM's Crash Meets Its Quantum Gambit
Transcript
- Today on Quickly Quantum: IBM had the worst single day in its history and turned around days later with its biggest quantum move in years — so is that a hedge or a panic move? Before that, in the headlines: PsiQuantum banks a hundred twenty-five million dollars from DARPA, neutral atoms had a genuinely great week, and a quiet zinc oxide discovery might matter more than any of the louder stories. Welcome back to Quickly Quantum, your daily brief on the quantum frontier. It's Saturday, July 25, 2026. Let's get into it. This was an IBM week wrapped inside a bigger quantum reality-check week. The stock crash and the HRL Laboratories acquisition sit right next to each other on the calendar, and the whole industry spent the week arguing whether hardware bets can justify sky-high valuations — even while national programs and funding rounds kept the money flowing in regardless. Here's the week that was. First, a story we covered earlier this week that's still rippling: IBM signing a definitive agreement to acquire HRL Laboratories, the Boeing and G-M-owned research lab famous for silicon-spin qubits — a totally different way of building a qubit than IBM's superconducting chips. The quick version: this gives IBM a second qubit modality to hedge its bets. Jay Gambetta told Reuters HRL has, in his words, "the strongest" spin qubit team in the world, and according to Reuters reporting, the technology mainly matters after IBM's Blue Jay system, which the company is targeting for around twenty thirty-three. Now, the timing here is what makes it interesting — this deal landed just days after IBM's stock cratered on earnings, and financial terms were never disclosed. So is this a genuine technical pivot, or a shiny distraction from a brutal quarter? Probably some of both. Let's get into exactly how brutal that quarter was, because that's the story underneath everything else this week. IBM's historic twenty-five percent stock crash reshapes its quantum narrative — and I want to spend real time on this one, because it's the backdrop for basically everything else in quantum this week. IBM suffered the worst single-day stock rout in its modern corporate history after a Q2 preliminary release missed the numbers badly: revenue came in at seventeen-point-two billion dollars against a consensus of seventeen-point-eight-six billion, earnings per share landed at two ninety-three against an expected three-oh-one, and mainframe and Z-series revenue — historically one of IBM's steadiest cash cows — was down roughly forty-two percent. All told, that erased roughly sixty-nine billion dollars in market value in a single session. Sixty-nine billion dollars gone in one trading session — for a company IBM's size, that's about as bad as an earnings day gets. Here's why it matters for quantum. When a legacy tech giant takes a hit like that, investors start asking where the growth story is. And IBM's answer, increasingly, is quantum. Reportedly, according to Susquehanna, IBM's quantum unit alone could be worth something near sixty-five dollars a share — around sixty-one billion dollars — which dwarfs the roughly one-point-one billion dollars in cumulative quantum contracts IBM has actually booked since twenty seventeen. Let that gap sink in: a sixty-one billion dollar valuation resting on a little over a billion in real, booked business. Analysts are pointing out that IBM's quantum valuation is built almost entirely on future promise, not current revenue — which is the exact same tension running through the broader NISQ-hype debate simmering across the field all year. IBM needs quantum to be the answer. Whether the market's buying it long-term is still an open question. Now, sticking with government money finding its way into unproven architectures — we covered this one earlier in the week too: DARPA committing a hundred twenty-five million dollars to validate PsiQuantum's photonic quantum computing designs, part of the agency's ongoing effort to stress-test underexplored systems for utility-scale quantum computing. On X, investor @EmmanuelInvest called it PsiQuantum's largest U.S. government award to date, and that tracks — DARPA validation money is a real vote of confidence for a company still working toward its first large-scale machine. But hold the applause for a second — validation funding is not the same as demonstrated, utility-scale performance. PsiQuantum hasn't shipped a working large-scale machine yet, and a hundred twenty-five million dollars buys scrutiny — the proof still has to come from working hardware. Now here's one that flew a little under the radar but might be the sleeper story of the week: a massive multi-institution roadmap for neutral-atom quantum computing — that's the platform where you trap individual atoms with lasers and use them as qubits, a real third contender alongside superconducting and trapped-ion systems. The author list reads like a who's-who: QuEra, Harvard, and dozens of collaborators, laying out a strategic plan for the full technology stack. On X, Pasqal put it plainly: "The neutral atom field just published its first shared roadmap. A strategic plan covering the full stack: qubit scaling and high-fidelity operations, quantum error correction, circuit compilation, distributed quantum computing through processor networking." The headline claim is that neutral atoms could reach what they call quantum utility within a decade — and interestingly, the bottleneck they flag isn't raw qubit count, it's the cost and complexity of the laser systems needed to control all those atoms precisely. And this lands during a week where neutral atoms genuinely had a moment — this is a platform that's been quietly building momentum for a while now, so there's real convergence happening around this modality right now. Here's my skepticism, though: the paper's own authors concede that building affordable, efficient laser control is still unsolved, and that's arguably the single hardest piece of scaling this platform up. Strategic roadmaps targeting a specific year — twenty thirty-two, in this case — also have a pretty poor historical hit rate across this entire industry. I'd call this a real signal about where smart people are converging their bets — just don't set a calendar reminder for twenty thirty-two. We also covered this earlier in the week: Northwestern University researchers demonstrating quantum entanglement distribution — that's the process of linking two particles so measuring one instantly tells you about the other — coexisting with live internet traffic on the very same fiber-optic cable. The quick version of why that matters: most quantum networking demos need a dedicated, quiet fiber line all to themselves. Doing it on a busy, shared line is a real step toward infrastructure you could actually deploy, rather than a pristine lab setup. It fits a broader theme running through the week — quantum-in-the-real-world demonstrations, alongside things like the Quantinuum-SoftBank white paper, all leaning into near-term, non-fault-tolerant value instead of distant promises. Now, the honest caveat here, and it's a big one: scaling a single-link demo like this into anything resembling a usable quantum internet is still the real hurdle, and nobody's cracked that yet. Here's a story that genuinely slipped past the weekday coverage and deserves a beat: a team led by Sungkyunkwan University, working with the University of Wisconsin-Madison and the University of Washington, identified an atomic defect inside zinc oxide — yes, the stuff in sunscreen and industrial coatings — with properties that look promising as a spin qubit, meaning a qubit built from the quantum spin of a single electron or defect inside a solid material. The timing here is almost too perfect. This is the same week IBM put spin qubits back in the spotlight with the HRL Laboratories acquisition, and this discovery shows the materials science pipeline behind that entire bet is still wide open — new candidate materials are still emerging rather than converging on one winner. On X, quantum commercialization strategist @drmikemyers flagged it simply: "World's first 'zinc oxide spin qubit' could advance scalable quantum devices." Now, let's be clear about what this actually is. This is an early-stage materials identification: researchers finding a promising atomic defect and characterizing its properties. Turning that into a working qubit device on a lab bench is a much longer project, likely years off, and integrating it into a real processor is further still. That's exactly why it's worth flagging now rather than later: the materials that end up mattering in five or ten years are being discovered in papers like this one, quietly, while everyone's watching the bigger corporate deals. Now let's talk about the sharpest distillation of the entire week's tension, and it comes from analyst commentary rather than a press release: should IBM break up? Should quantum get spun off into its own company? The thinking goes like this — after that historic stock crash, IBM's quantum division might actually be worth more to shareholders standing on its own than buried inside a wounded legacy hardware and software conglomerate still bleeding mainframe revenue. It's a bold thing to float days after a company announces it's doubling down on quantum with a whole new acquisition, but that's exactly the kind of contrarian question this week's numbers invite. I want to be really clear about what this is: pure analyst conjecture. IBM has given zero indication it's considering anything like this — no leaked memo, no boardroom source, nothing, just outside observers doing the math on that valuation gap we talked about a minute ago, and wondering out loud whether the math points somewhere IBM itself hasn't said. So, time for the Hype Check. Here's my case: the underlying tension is completely real — that gap between a sixty-one billion dollar quantum unit and a little over a billion in actual booked contracts isn't going away, and it's a legitimate strategic question for any board to eventually face. But there is exactly zero evidence IBM is entertaining a spin-off, and speculative pieces like this tend to get published every time a big company has a rough earnings week. The underlying tension here is real; the spin-off packaging is pure speculation. My rating: this one's a three. Last story before we close out the week: a concrete hardware jump in photonic quantum computing — that's the approach that uses particles of light instead of superconducting circuits or trapped atoms as qubits — that the weekday shows missed entirely. Researchers led by Hu and colleagues generated and measured a four-photon quantum state, meaning four individual photons linked together in a single quantum state, at a rate of twenty-seven hertz — that's twenty-seven times per second — on a single silicon nitride chip. And the headline number is efficiency: point-nine-four-three, meaning almost ninety-four-and-a-half percent of the photons they tried to generate and detect actually showed up in the measurement, which the researchers say beats prior silicon photonics implementations by more than a hundredfold. Why does that matter beyond the lab? Photon loss is basically the core enemy of photonic quantum computing — every photon that gets absorbed, scattered, or missed by a detector is information you just lost. So a hundredfold efficiency jump on a single chip is a genuinely big deal for a platform that's also getting real government money right now, with PsiQuantum's DARPA win landing the same week. The honest caveat, though: this is still a single-chip lab demonstration of a four-photon state. Getting from four photons on one chip to networked, error-corrected photonic processors doing useful work is an enormous distance — the kind that swallows a lot of promising early results in this field. Worth watching closely — it's a real lab result, and right now, that's all it is. So that's the week: IBM's worst day meeting IBM's biggest quantum swing, government money still flowing into unproven architectures, and quieter stories — zinc oxide, silicon nitride chips, a busy fiber line — doing real work under the noise. These are threads I'll keep pulling on next week, especially anything that clarifies whether IBM's quantum bet is a hedge or a headline. If this show's earning a spot in your weekend routine, hit follow wherever you're listening — it genuinely helps. 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!