Latest / Quickly Quantum / July 30, 2026 — IBM's Triple Quantum Advantage Claim Faces the Tracker
Transcript
- Did IBM just do what Google couldn't quite pull off back in twenty-nineteen — win the quantum advantage argument for good? That's the question hanging over the whole industry today, because IBM just rolled out not one, not two, but three separate quantum advantage claims in a single coordinated announcement, and every one of them is posted publicly for anyone to try to knock down. Before we get there, in the headlines: IonQ clears its last regulatory hurdle to close a billion-dollar chip foundry deal, a thread on X claims D-Wave is beating an eighty-GPU cluster on routing problems, and quantum stocks are having a rough month that IBM's big day won't fix by itself. Welcome back to Quickly Quantum, your daily brief on the quantum frontier. It's Thursday, July 30, 2026. Let's get into it. First up — IonQ just cleared the last regulatory hurdle standing between it and owning an actual chip factory. The company's one-point-eight-billion-dollar cash-and-stock acquisition of SkyWater Technology — the largest US-based semiconductor foundry — is set to close tomorrow, July 31st. This deal's been in motion since January, and now it's real. Why should you care if you don't own the stock? Because this is about supply chain independence. Quantum computers, especially IonQ's trapped-ion systems, need specialized chips, and right now a lot of that fabrication runs through third parties IonQ doesn't control. Owning SkyWater means IonQ controls its own chip pipeline end to end, from design to fab to finished processor. Now, IonQ isn't alone in this instinct — IBM has its own new foundry venture, called Anderon, running essentially the same calculus: if quantum hardware is going to scale, you don't want to be waiting in line behind Nvidia and everyone else for wafer capacity. It's a pattern worth noticing across the whole industry — hardware companies deciding that renting fabrication isn't good enough anymore, they want to own it outright. The interesting wrinkle is timing: this deal closes the same week IBM is out there claiming quantum advantage with its own hardware, so you've got the two biggest names in quantum computing making moves on completely different fronts — IBM on capability, IonQ on manufacturing control — inside the same forty-eight hours. Over on X, a thread from @NekiWeb3 got people talking about D-Wave's Advantage2 system — the claim is that it's already beating a cluster of eighty Nvidia H100 GPUs on a routing optimization problem. Caveat up front: this is one thread, we haven't independently confirmed that eighty-times number, and these benchmark claims usually depend heavily on which problem gets chosen to showcase. But the more interesting part isn't the speed claim, it's the pivot into trust. @NekiWeb3 argues the real bottleneck isn't whether quantum hardware can solve these problems, it's whether anyone can trust an external quantum computer with sensitive data, and points to a startup called Quip combining blind quantum computing — running a job without the operator ever seeing the actual problem — with blockchain-style zero-knowledge proofs to verify results. Replies in the thread split on whether that's practical for today's noisy hardware or overengineered for a problem that doesn't exist yet. Funny thing is, that's basically the same verification headache IBM is wrestling with today, just from the annealing side instead of the gate-model side. This next one lands in the exact same debate. A researcher named @qhrant, who's CEO of BlueQubit and holds a physics PhD from Stanford, posted a new arXiv preprint this week claiming quantum advantage for simulating something called prethermal dynamics — basically how a quantum system behaves before it settles into thermal equilibrium — and says it holds up on both an IBM superconducting processor and Quantinuum's Helios trapped-ion machine. On X, he wrote: quote, 'New paper with some amazing collaborators claiming quantum advantage for quantum simulation of prethermal dynamics on an IBM quantum processor and validated with Quantinuum's Helios. Try to break our classical hardness claim!' end quote. Worth noting BlueQubit is actually a named partner in IBM's own Quantum Advantage Tracker, so this claim lands directly inside the same live contest IBM opened today. It's a preprint, not peer-reviewed, from a player with skin in the game — call it reported, not settled. And this fits a pattern we've flagged before with Quantinuum specifically: milestone announcements that tend to arrive incrementally, with the scaling question always pushed to next time. Here's the reality check the industry always needs. IonQ, Rigetti, and D-Wave Quantum are all down roughly thirty percent over the past month, and Yahoo Finance is asking the obvious question: is more pain coming? Now, I don't have new numbers on where this bottoms out, but the timing here is almost too on the nose — the same week IBM is out declaring a new era for its quantum hardware, the three pure-play quantum stocks that live and die by exactly that kind of headline are getting hammered anyway. That's the gap between hardware milestones and investor confidence, and it's a gap that keeps showing up in this sector. Big technical wins from IBM don't automatically lift Rigetti or D-Wave, because those are different companies making different bets, but it does mean the market isn't just trading on vibes right now — actual capital is moving on something, and it's not enthusiasm. And last in the headlines — less flashy, but it matters if you care about where the money goes next. QED-C, an industry consortium, teamed up with the NSF-funded Center for Quantum Networks to publish a roadmap laying out what technology quantum networks actually need to get built — think the infrastructure equivalent of what fiber optic cable was for the regular internet, but for connecting quantum computers together securely. More than fifty experts contributed, including people from IonQ, L3Harris, and Aliro. This isn't a breakthrough, it's a planning document — but planning documents like this are how an entire industry decides where research dollars and grants flow for the next five years. If quantum networking ever becomes a real category, roadmaps like this one are where today's priorities get set down in writing. Our main story today: IBM's triple bet on quantum advantage — three separate claims, released together, all posted publicly where anyone can try to tear them down. Let's start with what actually happened, because there's a lot packed into today's announcement. IBM didn't put out one quantum advantage claim today, it put out three, each with a different partner, each targeting a different kind of problem. With Algorithmiq, IBM's team simulated something called heterogeneous matter — essentially modeling a system built from mismatched materials — on one of its Heron processors. With the University of Chicago, they encoded seventy logical qubits — logical qubits being error-corrected qubits built out of many noisy physical qubits stitched together so mistakes get caught and fixed — and used them to sample from a circuit that's supposed to be classically intractable, meaning no ordinary supercomputer can reproduce the answer in any reasonable amount of time, and they pulled it off in around fifteen minutes. And with error-mitigation partner Qedma, IBM modeled the dynamics of a seventy-four-qubit system and says it beat one of the most powerful supercomputers on the planet at that same task. All three results are sitting on IBM's open Quantum Advantage Tracker, which exists specifically so outside researchers can hunt for a faster classical algorithm that matches the quantum result — because if somebody finds one, the advantage disappears. Now, here's the context you need to actually feel why today matters. This is IBM CEO Arvind Krishna cashing a check he wrote back in April, when he told investors that partners would achieve, quote, 'the first examples of quantum advantage' this year using IBM hardware — part of a roadmap north of ten billion dollars aimed at a fault-tolerant machine called Starling by twenty twenty-nine. Fault-tolerant just means a computer that corrects its own errors well enough to run long, complicated calculations without noise wrecking the answer — that's the holy grail everyone in this field is racing toward. The University of Chicago piece is aimed squarely at the industry's oldest sore point: random circuit sampling, the technique where you run a circuit that should be hard for classical computers to simulate — and the problem has always been that once these circuits get complicated enough to be genuinely interesting, it gets brutally hard to verify the quantum computer actually did the job right instead of producing noise that just happens to look plausible. The paper is literally titled 'Sampling hard circuits with verifiably high fidelity,' and Bill Fefferman at UChicago frames the goal as building techniques that raise confidence you're solving something genuinely hard — not noise dressed up to look like an answer. That's the whole ballgame here: not just claiming advantage, but claiming it in a way skeptics can actually go check for themselves. So how worried should we be that this turns into another Sycamore moment? Quick rewind for anyone who wasn't around for it: back in twenty nineteen, Google claimed quantum supremacy — the predecessor term to quantum advantage — with its Sycamore chip, claiming a result no classical supercomputer could match in any practical timeframe. Within a couple of years, improved classical algorithms undercut that claim, and it didn't hold up nearly as cleanly as the headlines suggested. Now, here's the twist that should keep IBM humble today: IBM itself published one of the papers that helped undercut Google's claim back then. So more than almost anyone, IBM knows exactly how fast an advantage claim can age badly. Scott Aaronson, about as respected a voice as exists in quantum complexity theory, has made the point that these kinds of pronouncements are typically met with skepticism, and for good reason — classical hardness claims often rest on unproven complexity assumptions, meaning nobody has actually proven a classical computer can't eventually catch up, researchers have just failed to find the trick yet. That's not a knock on IBM specifically. Proving something is classically hard is genuinely difficult math, and the goalposts move every time somebody publishes a smarter classical algorithm. Which is exactly why the open tracker matters more than the press release does. Outside analysts are pretty blunt about it: the real test isn't today's announcement, it's whether the Quantum Advantage Tracker — the public, community-run site hosting these results — survives sustained attempts by outside researchers to find a classical shortcut. That's a genuinely different posture than twenty nineteen. Back then, Google's claim lived and died inside academic papers and press coverage, argued out after the fact. This time, IBM is inviting the takedown attempt in public, close to real time. Bill Fefferman's University of Chicago paper is built specifically to make that takedown attempt mean something, engineering in verification instead of just asking everyone to trust the result. And IBM's own camp isn't the only one wrestling with this question this week. That BlueQubit preprint on prethermal dynamics we just covered, and even the D-Wave and Quip conversation happening in a completely different corner of the field — annealing instead of gate-model hardware — are circling the identical question: can anyone outside the company that built the machine actually verify the result, not just marvel at it. Three different results, three different hardware families, all chasing that same unanswered question in the same week — that's a real signal about where this field's attention has shifted, not just a coincidence of timing. So where do I land on this? Time for the Hype Check. I'm putting this one at a seven out of ten on substance. Here's why: three separate demonstrations, posted publicly, with the University of Chicago piece specifically engineered so outsiders can try to break it — that's real, and it's a meaningfully different posture than past advantage claims that got announced and then left for the community to pick apart on its own timeline, months or years later. But it's a seven and not higher, because the track record here is genuinely bad. Google's claim didn't survive contact with better classical algorithms, and posting a result publicly isn't the same thing as independent researchers spending months actually trying and failing to break it — that clock just started today. The Qedma and Algorithmiq results in particular haven't had any real time to face that scrutiny yet. Ask me again in six months, once people have actually taken a real run at the classical shortcut, and this number could move in either direction. The open question I keep sitting with: if three advantage claims can launch on the same morning, what happens to the word advantage itself once the tenth one shows up — does it start meaning something specific and defensible, or does it just become the next quantum supremacy, a phrase everyone uses and nobody defines the same way? And if you're new here, today was a loaded one to jump in on — three quantum advantage claims, a billion-dollar foundry deal closing tomorrow, and a stock slump running in the opposite direction, all in one day. If you want that mix of headlines and honest analysis in your feed every weekday, 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!