Latest / Quickly Quantum / August 27, 2026 — IBM Buys a Rival Qubit's Home Lab
Transcript
- Today on Quickly Quantum: IBM just bought a lab built on a totally different kind of qubit than the one it's staking its whole roadmap on — so is Big Blue hedging its bet, or just buying insurance it never plans to use? Before that, in the headlines: Europe hands thirteen startups a shot at real EU financing, a physicist famous for hating quantum hype tells you to go read one anyway, and two companies pick a fight over silicon spin qubits on the very same day. Welcome back to Quickly Quantum, your daily brief on the quantum frontier. It's Thursday, August 27, 2026. Let's get into it. And if you thought the silicon-versus-superconducting fight was settled — think again. Now, first up: the EuroHPC Joint Undertaking — that's the European public body that pools funding for shared quantum and supercomputing infrastructure — just named the winners of its Quantum Grand Challenge. Thirteen European quantum-processor startups made the cut for Phase One funding, chosen out of twenty-seven eligible proposals from thirty submitted. Clear Phase Two, and each company could unlock up to thirty million euros in venture debt from the European Investment Bank. That sounds huge, and it could be, but it's a ceiling, not a check. The actual Phase One money on the table is modest — a total indicative budget of around four million euros split across all thirteen companies. So think of this less as Europe writing checks and more as Europe building a funnel, trying to turn grant-stage prototypes into companies investors will actually back later. Which thirteen startups actually make it to that second stage? That's the real story here, not the headline number — so if you're the type who tracks EU tech funding, this is the one to bookmark. Over at Lawrence Livermore National Lab and NIST, researchers just published a result in Physical Review Letters that sounds small but matters a lot if you're trying to build a trapped-ion quantum computer — that's the approach that traps individual charged atoms with electromagnetic fields and links them with laser pulses. The trouble with entangling those ions — linking their quantum states so an operation on one affects the other — is that outside a pristine lab, vibration and stray fields knock the fidelity down. This team found that ramping both the force and the detuning, easing into the entangling pulse instead of snapping it on, produced Bell-state fidelities above ninety-nine percent across a much wider range of conditions, including noisier setups. Bell-state fidelity, in plain terms, is how close to perfect that entangled link actually is. Now, this isn't a new qubit-count record, and it hasn't been independently replicated outside those two labs yet — it's an engineering fix for the gap between what works in a showcase and what works in a machine you'd actually run every day. And if you've ever wondered why trapped-ion computers still aren't running your workloads, this unglamorous fix is exactly the kind of thing that closes that gap. A physicist who's built a career calling out quantum hype just told her followers to go read a hype piece — that got my attention. Sabine Hossenfelder, who posts as @skdh on X, shared a Bloomberg feature this week on the tension running through this entire industry: quantum computers could deliver real breakthroughs in finance, medicine and climate modeling, and the same machines could eventually break the encryption the internet runs on. She called it, quote, an 'excellent article,' end quote — notable, because Hossenfelder doesn't usually hand out compliments to hype-adjacent coverage. The post drew dozens of replies debating exactly that split: promise versus threat. But why would a professional skeptic recommend a piece framed around a cybersecurity crisis? Now, I'll flag the same caveat I'd flag for anyone: the timelines for when a quantum computer could actually crack today's encryption are contested industry estimates, not settled fact, and this discourse is really about that framing catching on in the mainstream press, not a new technical claim. Still, you don't have to buy every quantum headline to find that framing useful — and when your loudest skeptic says the balanced take is worth your time, that's a signal in itself. In funding news out of Paris: ColibriTD, a quantum software startup, closed a four million euro seed round led by Earlybird Venture Capital, following an earlier one million euro pre-seed from the same firm. The money's going toward its H-DES platform — that stands for Hybrid Differential Equation Solver, a quantum algorithm built to solve the partial differential equations that govern fluid dynamics, structural mechanics and financial risk. ColibriTD's pitch is that hardware gets all the headlines in this industry, qubit counts and error rates, but software decides where the value actually lands — and they've already run seven co-construction projects with industrial partners including Thales, the European Space Agency and BMW. Four million euros is not a huge check by industry standards, and seven pilots is still pilots, not production deployments at scale. But if you've been wondering where quantum software money actually goes, this is a decent example of the bet: the algorithm layer, not just the chip, deciding where near-term value lands. And on the very same week IBM was closing a silicon-spin deal, Diraq — a company betting entirely on that qubit technology — opened a new twenty-person engineering hub in Santa Monica, focused on chip design, cryogenic control electronics and device modeling, with plans to double that headcount within a year. Quantum Computing Report is the one reporting this, and we haven't independently confirmed it elsewhere yet. Now, Diraq builds its silicon spin qubits — quantum bits based on the spin of a single electron in silicon — using standard commercial chip-manufacturing processes, which is the whole argument for why this modality might scale cheaper than exotic alternatives. Two separate companies, two separate decisions in the same stretch of days, both pointing the same direction: silicon spin just went from research curiosity to something worth building real infrastructure around, worth noting if you're mapping which qubit types are actually attracting money right now. Which brings us to the story that made that connection unavoidable. Our main story today: IBM has officially completed its acquisition of HRL Laboratories — and the thesis I want to walk you through is whether Big Blue is quietly hedging its own quantum bet, or just buying insurance on a modality it insists it will never actually switch to. HRL was, until now, jointly owned by Boeing and General Motors — a research institution with decades of work in silicon-spin qubits, quantum sensing, materials science, cryogenics, control electronics and chip packaging. Both Boeing and General Motors will keep partnering with IBM and HRL on quantum applications going forward, according to IBM's own announcement. Now, IBM has spent years telling anyone who'll listen that superconducting qubits — tiny loops of current cooled to near absolute zero — are its path to fault tolerance, the point where a quantum computer catches and corrects its own errors faster than they pile up. Its roadmap runs through a machine called Starling, targeted for 2029 and meant to perform a hundred million quantum operations, followed by a second machine, Blue Jay, aimed at the mid-2030s. So here's the tension: IBM just bought a lab whose core expertise is a completely different way of building a qubit. IBM's own materials call this complementary, not competitive — folding HRL's silicon-spin and packaging know-how into a broader hardware push that could also feed Anderon, IBM's own quantum wafer foundry. Financial terms of the deal were not disclosed, which makes it genuinely hard to gauge how serious a bet this is versus a modest research add-on. We've talked before on this show about IBM needing quantum to be the answer to bigger strategic questions the market's been asking about the company — this deal is worth weighing against that. So does this actually mean IBM has doubts about superconducting qubits scaling on its own published timeline? I don't think the acquisition alone proves that — you don't buy a lab's patents and people because you've lost faith, necessarily. But you also don't buy a rival modality's deepest expertise if you're a hundred percent certain your own path gets you all the way to fault tolerance alone. The honest answer is we can't fully know, because IBM didn't disclose what it paid, and it hasn't said how fast HRL's silicon-spin work gets folded in versus run as its own separate track. That ambiguity is doing a lot of work here, and I'd rather say that plainly than pretend the press release settles it. Now, what I can tell you is the context around it is not subtle: this is the same week Diraq opened a new U.S. engineering hub explicitly to scale silicon spin qubits, and EuroHPC just funded a batch of processor startups working across multiple qubit types. Silicon spin, a modality everyone respected but nobody funded like a contender for years, is suddenly pulling in serious industrial money from more than one direction at once. Does IBM end up running two real hardware programs, or does it just quietly bank HRL's patents and packaging know-how? That's the open question, and it's the same one we've been asking about IBM's broader quantum bet for a while now — whether the company actually believes in this, or just needs to look like it does. This deal doesn't resolve that. It raises the stakes on the answer. Time for the Hype Check. I'm calling this one a five. Real acquisition, real capabilities changing hands — silicon spin qubits, cryogenics, packaging, actual engineers — but with no disclosed price tag and no committed integration timeline, there's no way yet to tell whether this is a genuine hedge or a tidy line in a press release. If HRL's silicon-spin work shows up inside an actual IBM roadmap update by the time Starling ships in 2029, that tells us this was the real hedge it looks like today. Now, if you're new to the show, follow Quickly Quantum wherever you listen so tomorrow's episode finds you automatically. 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!