Latest / Quickly Quantum / August 1, 2026 — Weekly Recap: Does IBM's Advantage Claim Actually Hold Up?
Transcript
- Today on Quickly Quantum: IBM says it's cracked quantum advantage for real this time — verified, logical results a classical computer literally can't check. But this same week produced its sharpest skeptic in years, arguing the whole scaling dream might be running straight into a wall of noise. Before we get into that fight, here's the shape of the week: IonQ closed a chip-manufacturing acquisition even as its stock — along with Rigetti and D-Wave — kept sliding, Multiverse Computing chased one of the biggest raises in the sector, and a handful of quieter stories on spin qubits, encryption, and Navy research strategy flew under the radar. Welcome back to Quickly Quantum, your daily brief on the quantum frontier. It's Saturday, August 1, 2026. This is our weekly recap — the stories you caught pieces of, tied together, plus what you missed. Let's get into it. IBM's big swing landed this week, and the short version goes like this: IBM and the University of Chicago say they've cleared a real bar for quantum advantage — not just speed on some synthetic puzzle, but a computation verified as correct that no classical machine could check on its own. The team ran seventy logical qubits, that's error-corrected qubits built from many noisy physical ones working together, through two thousand four hundred fifteen logical two-qubit operations and four hundred sixty-eight logical T-gates. IBM Research Director Jay Gambetta didn't hedge at all: 'We are now firmly in the quantum advantage era.' Here's the catch worth sitting with, though — IBM defined this advantage on its own benchmark, and by its own account dared the community to prove it wrong, which means independent replication hasn't actually happened yet. That's not a small caveat — 'verified' here means IBM's own team checked its own math, not that an outside lab reran the experiment and got the same answer. We've said before on this show that IBM needs quantum to be the answer, and whether the market's buying that long-term is still an open question. This week's result is IBM's strongest pitch yet toward that goal. It just isn't outside confirmation. IonQ actually closed something this week instead of just teasing it: the company cleared its final regulatory hurdle and completed its acquisition of SkyWater Technology, giving IonQ a US-based chip manufacturing base instead of renting capacity from outside foundries. IonQ's own account marked the moment on X, casting the completed deal as a strategic win for its enterprise and government partners. That post pulled in easily the loudest engagement in our commentary pile this week. Now, the timing here is worth sitting with — this deal-making enthusiasm lands the exact same week IonQ's stock, along with Rigetti and D-Wave, kept sliding hard. So the real question isn't whether vertical integration is smart strategy long-term. It's whether investors are willing to sit through the slide to find out. Here's the number that undercuts both of those stories: IonQ, Rigetti, and D-Wave are all down roughly thirty percent over the past month, even as their companies keep posting technical milestones like the SkyWater deal we just covered. That's the tension of this entire week in one sentence — the science headlines keep climbing, and the stock charts keep falling. Now, to be fair, pure-play quantum stocks are small and volatile by nature, and a thirty percent swing in a month isn't necessarily a verdict on the underlying technology — it happens to speculative small-caps all the time. But it's also a reminder that 'quantum advantage' as a phrase and quantum advantage as a revenue line are two very different things, and right now the market seems a lot more interested in the second one. If you're tracking whether the hype and the money are moving in the same direction, this month says no — and that gap matters every single time one of these companies drops a press release. Money kept moving on the private side too, even while public stocks wobbled. Multiverse Computing announced it's targeting up to five hundred seventy million dollars in a Series C round — one of the largest funding pursuits anywhere in the sector this year. It's part of a broader pattern this week of capital still flowing into quantum even as the public markets we just talked about kept sliding. Now, a target isn't a close — announcing you're pursuing five hundred seventy million dollars doesn't guarantee the round lands anywhere near that number, and we'll want to see final terms before calling this one done. Series C rounds this large in quantum are still rare enough that even a partial close would be notable on its own. But directionally, this tells you something: private investors with longer time horizons than public-market day traders are still writing checks into this space, even as retail sentiment on the listed names sours. That split between private conviction and public skepticism might be the most honest snapshot of where quantum computing actually stands right now. Then there's the counter-programming nobody scheduled, and it landed today. Gil Kalai, a mathematician at Hebrew University and Reichman University, sat down with Yuval Boger for a podcast interview laying out his long-standing skepticism of scalable quantum computing, built on two arguments. First, correlated noise — errors that don't stay isolated and independent the way error-correction theory generally assumes, but instead show up linked across qubits in ways that could defeat the whole correction scheme. Second, a complexity-based argument that today's noisy, non-error-corrected devices — what the field calls NISQ machines — may be fundamentally incapable of ever reaching genuine quantum supremacy, no matter how many qubits you add. Now, Kalai isn't some fringe figure shouting into the void — a thread from @HeidingOut on X this week made almost exactly that point, listing him alongside other longtime skeptics like Mikhail Dyakonov, Robert Alicki, and Leonid Levin as scientists the field's boosters tend to talk past rather than engage with. But Kalai has real critics too. Scott Aaronson has argued that Kalai's position starts from the conclusion — quantum computing is impossible — and works backward to justify it, rather than following the evidence wherever it leads. That's the honest shape of this fight: two serious thinkers, reading the same noisy hardware, and reaching opposite conclusions about whether the noise ever actually goes away. So here's why this lands where it does: the same week IBM stood up and said it's 'firmly in the quantum advantage era,' the field's most persistent skeptic was making the case, with real technical substance, for why that era might be an illusion built on noise nobody's fully priced in. Neither side has definitively won that argument. That's just honestly where the field stands. Away from the headlines entirely, two independent research groups published studies this week tackling one of the least glamorous but most important problems in the field: wiring. Semiconductor spin qubits — quantum bits built using roughly the same manufacturing techniques as ordinary computer chips — are one of the most promising paths to scale, mainly because that manufacturing familiarity means you could theoretically mass-produce them using processes the chip industry already knows cold. The catch has always been control: connecting distant qubits to each other, and controlling huge numbers of them at once, without needing an unmanageable tangle of wires running in and out of a dilution refrigerator. Think of it this way — adding more qubits is the easy headline number. Wiring each one up without turning the whole machine into a rat's nest of cables is the actual engineering fight, and it's the fight that decides whether any of these architectures scale past a few dozen qubits into the thousands. These two new studies together made real progress on exactly that bottleneck — connecting qubits that are physically far apart, and controlling large numbers of them without that unmanageable tangle of wires. Neither is a finished, scaled architecture. These are lab-stage demonstrations, and turning a wiring fix into a working large-scale chip is still years of engineering away. But this is exactly the kind of story that gets buried under IBM press releases and stock-slide headlines, even though it's arguably more load-bearing for the field's long-term future than either of those. Superconducting qubits, the kind IBM and Google build with, get almost all the public attention. Spin qubits are quieter, cheaper to fabricate at scale, and this week, out of view of most of the coverage, they took two genuine steps forward. On the policy side, NIST — the National Institute of Standards and Technology — finalized three algorithms for post-quantum cryptography, encryption methods built to survive an attack from a quantum computer powerful enough to break the math underlying today's internet security. For a non-technical listener, here's why that matters: data intercepted and stored today could be decrypted retroactively once a quantum computer capable enough exists — the so-called 'harvest now, decrypt later' threat — so the migration clock on this stuff effectively started years before the quantum computer that breaks anything actually shows up. This finalization is the missing anchor for a broader crypto-agility conversation happening across finance and government right now, meaning building systems that can swap encryption methods without a total rebuild. Now there's an actual finalized standard to point to. Here's the honest caveat, though, and it's the same one that applies every time a standards body finishes its work: finalizing an algorithm on paper is not the same as migrating the world's actual infrastructure to use it. Banks, governments, and cloud providers still have to rip out or wrap decades of existing cryptographic plumbing, and that migration — not the math — is where these efforts tend to slow down or stall entirely. So mark this as real progress on the defensive side of the quantum story, the side that doesn't get nearly the coverage 'quantum advantage' claims do, but treat the finalized standard as the start of a much longer adoption clock, not the finish line. The Navy's quantum arm quietly put out its own strategy document this week. The U.S. Naval Research Laboratory detailed the operational roadmap for its Quantum Science Institute, laying out how its research lines up with national quantum executive-order directives across three areas: sensing, computing, and networking. Sensing covers things like ultra-precise navigation and detection that don't depend on GPS; networking covers the quantum-linked communication infrastructure meant to eventually connect quantum computers and secure channels together. It's the kind of story that never makes a splash on its own — no dramatic qubit count, no stock ticker attached — but it fits a pattern we've seen elsewhere this month, where different governments are formalizing their own quantum strategies in parallel. Israel, for instance, has been pushing for its own national quantum research center around the same time. The crypto-agility push we just talked about is one thread in the U.S. version of that; this Naval Research Lab roadmap is another. Now, the honest read on any strategic-priorities document is the same every time: publishing a roadmap is the easy part. Concrete funded programs, actual procurement contracts, and hard timelines are the real test of whether an institution follows through on paper priorities, and those details are exactly what a document like this tends to gesture at without locking down. Still, worth noting in a week dominated by corporate press releases: quiet, methodical government planning work like this is happening in parallel, and it's usually a better long-run predictor of where public money in this field actually goes than any single company's advantage claim. So here's the marker I'm setting heading into next week: if an independent team actually replicates IBM's seventy-logical-qubit result — not just cites it, replicates it — that tells us this advantage-era talk has real legs. If months go by and nobody outside IBM can reproduce it, Gil Kalai's noise argument gets a lot more interesting than a lot of people want to admit right now. If this show's useful to you, follow it wherever you're listening so next week's verdict lands in your feed 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!