Latest / Quickly Quantum / July 23, 2026 — IBM's $10B Quantum Bet Adds HRL Labs
Transcript
- Today on Quickly Quantum: IBM just went shopping — and the target might tell you more about where quantum computing is really headed than anything the company's own superconducting chips have shown us yet. Before that, in the headlines: PsiQuantum lands its biggest U.S. government contract ever, Japan joins the silicon-qubit race alongside Intel, and a Chicago bank finds a real hint of quantum speedup in routing packages. And if you've been assuming every quantum company sticks to one flavor of qubit — buckle up, because today's big story says otherwise. Welcome back to Quickly Quantum, your daily brief on the quantum frontier. It's Thursday, July 23, 2026. Let's get into it. Now, first up — big number, and it's real: PsiQuantum just signed an expanded one hundred twenty five million dollar agreement with DARPA, the Pentagon's advanced research arm, under the Quantum Benchmarking Initiative — think of it as the government's toughest audit of which quantum hardware actually works. That's PsiQuantum's largest U.S. government award to date, and it makes them one of only two companies to reach the program's final validation stage, called Stage C. This builds on a thirty one point eight million dollar Stage C award from last September and a separate hundred million dollar CHIPS Act letter of intent. CEO Victor Peng told Reuters the vetting will help the industry figure out, in his words, what's real and what's not, and he says the company's first commercial-scale machine should be running around twenty thirty. Worth remembering: this is validation testing, not a finished machine — the utility-scale system is still years out by Peng's own timeline. On X, PsiQuantum's own account called it, quote, our company's most valuable government award to date. Chief of Staff Alex Mack, the handle @alexmmmack, put it more casually: quote, just another week at PsiQuantum. And Crain's Chicago Business ran it straight, just the headline: PsiQuantum gets one twenty five million from DARPA. Now, over in Japan: Hitachi has launched a government-backed research project with Intel K.K. — that's Intel's Japan entity — and AIST, the country's national industrial science institute, to advance silicon-based quantum computing. Silicon qubits, sometimes called spin qubits, store quantum information in the spin of an electron sitting inside ordinary semiconductor material, which matters because it means leaning on decades of chip-manufacturing know-how instead of inventing an entirely new fabrication process from scratch. This adds Japan to a growing list of governments betting on silicon as a scalable alternative to the superconducting circuits IBM and Google use, or the trapped-ion approach IonQ favors. The bet is straightforward: if silicon qubits can be made to work reliably, the existing semiconductor industry already knows how to manufacture them at scale, which could shortcut years of custom tooling other approaches still need to build from zero. Now, here's a nice one for the quantum internet file: researchers at Northwestern University showed that fragile quantum entanglement signals — pairs of particles linked so measuring one instantly tells you about the other — can travel over the same busy telecom fiber-optic cable that's simultaneously carrying your regular internet traffic. Why does that matter? Because a lot of quantum networking demonstrations up to now needed dedicated dark fiber — cable laid specifically for the experiment, sitting empty of any other signal. If quantum and classical data can share the same glass that already crisscrosses every city, you skip the enormous problem of laying brand-new infrastructure just to build a quantum network. Phys.org covered the result independently too, which is a good sign this isn't just a university press release running unchecked. It's one of those unglamorous results that won't make headlines the way a new qubit record does, but it's exactly the kind of engineering that decides whether quantum networking ever leaves the lab. Now, over to China: Shanghai-based Taiyi Quantum raised forty two million dollars to advance its neutral-atom quantum computing platform — the approach that traps individual atoms in place with laser light and uses them as qubits, the same general technique companies like Infleqtion, QuEra, and Atom Computing are chasing here in the U.S. We're hearing this one from The Quantum Insider and haven't independently confirmed the figure elsewhere yet, so take the number with that caveat. But the bigger picture holds regardless of the exact total: China keeps pouring private and state money into neutral-atom hardware, a modality a lot of Western quantum executives have quietly started treating as a strong near-term bet, precisely because it scales without needing the dilution refrigerators superconducting chips require. Worth watching whether Taiyi's roadmap actually lines up with the more mature neutral-atom players, or whether this is just capital chasing an increasingly crowded field. Now, Infleqtion — the neutral-atom company out of Colorado — says it'll deploy a fault-tolerant quantum computer in Illinois designed to demonstrate more than fifty logical qubits, with a target of one hundred, built from more than a thousand physical qubits underneath. Quick gloss: a logical qubit is an error-corrected qubit, stitched together from many imperfect physical qubits so the errors cancel out — it's the unit that actually matters for running real algorithms, and the whole industry is racing to get useful numbers of them. Infleqtion is pitching this as a center of excellence for applying quantum computing to the American energy grid, and it comes alongside three new Department of Energy Genesis Mission projects the company just landed for quantum computing and sensing research. The caveat matters here: fifty and a hundred logical qubits are roadmap targets, not achieved numbers — Infleqtion hasn't built this machine yet, they've announced where they're building it and what they're aiming for. Now, over in Brussels: Europe's Quantum Flagship, the EU's long-running quantum research program, is lobbying the European Commission to ring-fence a dedicated three point five billion euro quantum budget in the bloc's next multi-year funding framework, rather than lumping it into a shared pot with other digital technologies. The worry, according to the Flagship, is that shared digital-tech funding gets diluted, and Europe can't out-compete the kind of concentrated state investment the U.S. and China are putting behind quantum. The Commission's public consultation on a European Quantum Initiative is open through October sixteenth, so this is a lobbying position going into that process, not a confirmed allocation — the Commission hasn't committed to ring-fencing anything yet. Still, the number itself is a signal: three and a half billion euros is what Europe's own quantum community thinks it needs just to stay in the race. Now, last quick hit, and it's a fun one for the optimization nerds: JPMorganChase, working with Argonne National Laboratory, found that QAOA — the Quantum Approximate Optimization Algorithm, a near-term algorithm designed to run on today's noisy quantum hardware — may deliver real speedups on complex logistics problems, like routing ten thousand packages efficiently. That's a genuinely interesting data point in a debate that's run for years: can these near-term algorithms actually beat classical heuristics on real-world optimization, the kind finance and logistics companies have long eyed as an early practical use case? Worth being careful here, though — QAOA speedup claims have been challenged by classical algorithm improvements before, more than once, so this is promising research, not a settled case of quantum advantage. Still, it's the kind of incremental, credible result that adds up over time — and speaking of companies making moves that add up, let's get to today's big one: IBM just went and bought a piece of Boeing and General Motors. Main story today, and I'm calling this one IBM's insurance policy, because that's really what this acquisition looks like once you get past the press release. IBM announced it will acquire HRL Laboratories, the Malibu, California research lab jointly owned by Boeing and General Motors, and if you don't know HRL by name, you've probably heard its work — it's built a reputation as a world leader in solid-state spin qubits and quantum networking, two technical areas that sit outside IBM's own core bet. Let's ground this in plain English first. IBM's entire quantum computing business is built on superconducting qubits — tiny circuits cooled to near absolute zero that behave quantum-mechanically. It's the same basic technology Google uses, and it's the platform IBM has staked its whole public roadmap on: a fault-tolerant, error-corrected quantum computer — meaning one that can run long, complex calculations without noise wrecking the answer — by twenty twenty nine. HRL, on the other hand, works in a different lane entirely: silicon and silicon-germanium spin qubits, which store quantum information in the spin of individual electrons inside chip-like material, plus quantum networking work that includes entanglement-swapping experiments with satellites, done alongside Boeing. HRL also built spinQICK, an open-source toolkit for the cryogenic control electronics — the chip-level hardware that talks to qubits at extremely cold temperatures. Now, why does this land now specifically? Because it's not IBM's first quantum announcement of the year. Back in June, the company disclosed it's committing more than ten billion dollars to quantum computing over five years, and it explicitly said that money would go toward research, manufacturing, ecosystem partnerships, and — this is the relevant word — acquisitions, all tied to that twenty twenty nine fault-tolerant target. So this HRL deal isn't a surprise pivot, it's IBM cashing a check it already told investors it was going to write. Here's the part that actually matters if you're trying to read the tea leaves on where quantum computing is heading: IBM has spent years as one of the loudest voices insisting superconducting qubits are the winning path. Buying a lab whose signature expertise is a completely different qubit technology — plus manufacturing know-how in silicon fabrication IBM doesn't have in-house — is IBM quietly admitting that betting everything on one modality might not be the safest move, or at minimum, that owning optionality on an alternative is worth the price of admission. And that's before you even get to the quantum networking piece, the satellite entanglement-swapping work with Boeing, which points at IBM wanting a foothold in linking quantum computers together over distance, not just building bigger single machines. Now, the caveat, and it's an important one: this deal was announced through IBM's own newsroom, and at the time of this reporting there's no independent, third-party confirmation of the actual deal terms — what IBM paid, how many of HRL's researchers stay on, or what the integration timeline looks like. It fits logically with IBM's own June disclosure about earmarking M&A dollars, so the shape of the story checks out. But until we see confirmation from outside IBM's own press shop, the specifics here are IBM's framing of IBM's deal. So let's dig into the deeper read, because there's more going on than just IBM buying a lab. First, think about the pedigree here. HRL has spent decades doing serious, publication-grade research for two of the largest industrial companies in America, Boeing and General Motors, which means IBM isn't just buying patents and lab equipment, it's buying a research culture and a track record that's survived scrutiny from aerospace and automotive engineering standards — a different kind of rigor than a typical quantum startup pitch deck has to survive. Second, the manufacturing angle is underrated. Silicon and silicon-germanium spin qubits are attractive precisely because they can, in theory, ride the existing semiconductor manufacturing base — the same fabs that make your phone's chip could, with enough engineering, make quantum chips too. That's the same logic behind the Hitachi-Intel-AIST silicon project we just covered, and it's the same logic drawing governments in the U.S., Japan, and the Netherlands toward silicon qubits as the modality most compatible with industrial-scale production. If IBM's serious about owning silicon spin-qubit fabrication know-how, that's a different kind of asset than another patent portfolio — it's a manufacturing shortcut. Third, and here's where I want to push back a little: announcing an acquisition through your own newsroom, with no price tag, no headcount numbers, and no integration timeline, is a very specific kind of disclosure. It's the corporate equivalent of a press release doing exactly what a press release is designed to do — shape the narrative before anyone else gets to. That doesn't make the deal fake, and it fits cleanly with IBM's own June commitment to M&A dollars, so I believe the deal is real. But 'IBM says it acquired HRL' and 'we know what IBM actually paid and what IBM actually gets' are two different levels of confidence, and right now we're only entitled to the first one. What would move me from cautiously interested to genuinely convinced? Three things. One, an actual dollar figure — even a rough one — confirmed outside IBM's own channels, whether that's an SEC filing, a Boeing or GM disclosure, or independent reporting. Two, clarity on whether HRL's spin-qubit and networking teams stay intact as a unit or get folded into IBM's existing quantum organization, because these lab acquisitions quietly lose their best people all the time. And three, whether IBM actually ships anything using HRL's technology, or whether this ends up as a defensive purchase that never surfaces in a real product roadmap. Time for the Hype Check. I'm putting this one at a six out of ten on substance. The strategic logic is completely sound — IBM diversifying beyond superconducting qubits, picking up real manufacturing and networking expertise, timed exactly to money it already told investors it would spend. That's a coherent story, not vaporware. But the actual deal — the price, the people, the integration — exists right now only in IBM's own words, and a company describing its own acquisition in the most favorable light possible is not new information, that's just Tuesday. Until there's outside confirmation of the terms, I'm filing this as real move, unverified specifics, and that's exactly the gap between a six and something higher. These are threads worth pulling on through the rest of the year — watch for any SEC filing from IBM, Boeing, or GM that actually puts a number on this deal, because that's the tell on whether the ten billion dollar quantum bet is being spent wisely or just spent quickly. If the show's earning your time, hit follow wherever you're listening, and if you've got a colleague who still thinks quantum computing is science fiction, this is a good episode to send them. 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!