Multiverse's $570M Round: Quantum Cred or AI Hype?

Five hundred seventy million dollars — that's the size of the check a quantum-rooted AI company just cashed, one of the largest funding rounds this industry has ever seen. Today on Quickly Quantum: Multiverse Computing's new billion-dollar valuation, and whether it says anything at all about quantum computing, or everything about how investors are chasing the word 'quantum' on what's fundamentally a classical AI compression business. Before that, in the headlines: a thirty-second coherence time out of trapped ytterbium ions, IonQ's cat-state trick that speeds up error correction three-fold, a federal deadline pushing contractors toward post-quantum encryption, and a theoretical result suggesting more qubits could mean your computation freezes solid instead of running faster. Welcome back to Quickly Quantum, your daily brief on the quantum frontier. It's Monday, July 27, 2026. Let's get into it.

Now, let's start with some genuinely good physics news. Researchers working with trapped ytterbium ions — a favorite atom for both quantum computers and atomic clocks — found certain quantum states that stay stable for more than thirty seconds. That might not sound like much, but in the quantum world, where information usually leaks away in milliseconds, thirty seconds is an eternity. This matters because coherence time — how long a qubit holds its information before noise scrambles it — is one of the biggest levers on error rates in trapped-ion quantum computers, the same technology IonQ builds its machines around. Longer coherence means fewer errors creep in, which means less overhead spent correcting them. The result comes independently from The Quantum Insider and Quantum Zeitgeist, both citing work out of the Universities of Amsterdam and New South Wales — solid corroboration. The caveat: this is a clean lab result on individual ions, not yet built into a working multi-qubit system, so how it translates into real fidelity gains at scale is still an open question. But it's exactly the kind of foundational number that tends to resurface in someone's roadmap a few years down the line.

Now, sticking with IonQ for a second — a new paper out of the company, posted to arXiv on July seventeenth by Webster and Delfosse, claims a real speedup on one of the hardest parts of fault-tolerant quantum computing. It's a cat-state-based protocol — cat states are a way of encoding a qubit so errors show up as detectable jumps rather than silent corruption — and pairing that with qLDPC codes, low-density parity-check codes borrowed from classical communications that use far fewer physical qubits than the standard surface code, they get roughly a three-times speedup on joint logical measurements, and up to seventy-four times in some circuits. This runs on simulations of IonQ's Q70 and Q102 walking-cat architecture, part of the company's broader fault-tolerant blueprint. We checked the paper directly on arXiv, number 2607.16166, and the math holds up. The catch: it's simulation, not physical hardware execution, and it hasn't seen much mainstream trade-press pickup yet. But if you're tracking how IonQ plans to get from today's noisy qubits to error-corrected ones without needing a forest of extra physical qubits, this is a meaningful data point.

Now, a policy story with real teeth. Security vendor Securosys says an executive order — Executive Order 14412 — has effectively set a 2030 deadline for U.S. government contractors to adopt NIST-certified post-quantum cryptography, the new encryption standards designed to survive a future quantum computer capable of breaking today's codes. Securosys says its own hardware security modules already support the certified algorithms, so if you're a contractor, the clock's now officially running. This fits into a bigger Washington push this year — quantum funding out of the Commerce Department, executive orders setting agency timelines — all aimed at the 'harvest now, decrypt later' threat, where adversaries steal encrypted data today and just wait for the quantum computer that can crack it. Now, worth flagging: this story comes from Quantum Zeitgeist citing Securosys itself, a vendor with an obvious commercial stake in selling post-quantum hardware, and we haven't independently confirmed the deadline framing beyond that one source. Take the urgency with a grain of salt — but the underlying trend, contractors needing to migrate before 2030, tracks with everything else we've seen out of Washington this year.

Now here's one that's been making the rounds on X. A widely-shared breakdown of a talk by IonQ engineer Daniel Pompa — posted by the account @TechInnovationz — argues that IonQ's laser-free trapped-ion technology, inherited from its Oxford Ionics acquisition, can be built inside standard semiconductor fabs. The thread says IonQ has already taped out a 256-qubit chip and claims a path to two million qubits by 2030 through its SkyWater fab acquisition. The framing, as @TechInnovationz relays it: trapped-ion scaling is now an engineering problem, not a physics one. That's a direct jab at the strategy IBM and Google have taken instead — hedging across multiple qubit types because nobody's sure yet which one wins. And that's a tension we've flagged on this show before: IBM needs quantum to be the answer to its own bet, and whether the market's buying that long-term is still very much an open question. Now, the honest caveat here — this is a company roadmap relayed through a fan account, not independently audited, and qubit-count roadmaps across this entire industry have a rough track record of actually hitting their stated dates. File it as ambitious, not confirmed.

Now, a bit of a buzzkill for the scaling-up crowd — and I mean that as a compliment to good science. Researchers at HZDR, the Helmholtz-Zentrum Dresden-Rossendorf, found that frequent disruptions can trigger the quantum Zeno effect, where repeated tiny disturbances act like constant measurements and can nearly freeze a quantum computation altogether. As @CyberWatch05 put it on X, quote, 'Quantum Zeno effect could freeze computations as qubit systems scale up,' end quote — going on to note that in adiabatic quantum computers, even tiny environmental disturbances can act like constant measurements, slowing or nearly halting computation as qubit counts grow. That's the sting: the effect gets worse, not better, as you add more qubits, cutting against the usual more-qubits-more-progress narrative. This is theoretical, drawn from adiabatic quantum computing models, not yet tested at real hardware scale — a caution flag, not a verdict. But it's a distinct failure mode from ordinary decoherence, and it deserves a spot on scaling roadmaps, not a footnote.

Now, mark your calendar: IonQ says it'll report second-quarter results on August fifth. That date's become a bit of a lightning rod — IonQ became the first quantum computing company to top a hundred million dollars in annual GAAP revenue this year, and bearish pieces, including some from The Motley Fool, have been circling the stock's valuation ahead of the report. Here's an interesting wrinkle @quantum_nyang flagged on X: quote, '$IONQ's customers usually surface only inside IonQ's own filings. On Aug 4, one of them opens its books first. Horizon Quantum (Nasdaq: HQ) — which agreed to buy IonQ's 6th-gen 256-qubit system — reports Q2 results Aug 4, before the open. IonQ reports the next day, Aug 5,' end quote. That's a genuinely useful tell — a rare chance to see how a real customer is talking about its IonQ purchase before IonQ itself gets to frame the story. Worth watching both dates back to back.

And finally in the quick hits: BlueQubit, working with Microsoft, Argonne National Lab, Sandia, and several universities, landed one-point-five million dollars from the Department of Energy's Genesis Mission program to apply AI to quantum error correction — using machine learning to design better error-correcting codes and speed up decoding, the process of figuring out what went wrong in real time so a system can fix it before the answer's ruined. As @IntEngineering put it on X, quote, 'BlueQubit is using AI to solve one of quantum computing's biggest problems,' end quote. It's part of a broader pattern — IBM has its Qiskit Orbit tool, IonQ has its own decoder work — everybody's betting AI can chip away at the physical-qubit overhead and latency that error correction creates. Let's be clear-eyed, though: a million and a half dollars is a research grant, not a result. It funds the attempt; it doesn't announce a breakthrough. Now, that word — AI — is about to show up again in a much bigger way, because our main story today is about the biggest funding round we've seen out of anything with 'quantum' in its name.

Now, let's get to our main story. I'm calling it 'quantum's money problem — or is it AI's?' because by the end of this segment I want you deciding which one it actually is. Here's the headline: a company called Multiverse Computing, based in San Sebastián, Spain, just closed a Series C funding round of five hundred seventy million dollars — about five hundred million euros — at a one-point-seven billion dollar pre-money valuation. Pre-money means before this new cash even lands, the company was already being valued at one-point-seven billion, and that's a five-times step-up from its Series B, which closed just over a year ago, in June of twenty twenty-five. This is, by any measure, one of the largest funding rounds any company with 'quantum' in its story has ever raised. It brings Multiverse's total funding to eight hundred million dollars. Who's writing these checks? The round was co-led by Forgepoint Capital International, BNP Paribas's Solar Impulse Venture Fund, and Bullhound Capital, with participation from a genuinely sprawling list — Qatar Development Bank, HP, Orange Ventures, Santander Alternative Investments, Tikehau Capital, Scania Invest, National Bank of Canada's venture arm, and a handful of European public and regional funds. JP Morgan and Santander advised on the deal. This isn't a scrappy seed round with two angel investors — this is institutional, sovereign-adjacent money, at serious scale. So what does Multiverse actually do? This is where the 'quantum' label needs some unpacking, because Multiverse isn't building a quantum computer, and it isn't selling algorithms to run on one. Its product is called CompactifAI, and what it does is compress large AI models — the kind that normally need a data center full of GPUs to run — by eighty to ninety-five percent, with what the company describes as minimal accuracy loss. Shrink a model that much, and suddenly it can run on a phone, or a factory floor with no cloud connection, instead of needing to phone home to a hyperscaler every time it thinks. Here's the quantum connection, and it's a real one, not just branding: the compression technique is built on tensor networks — a mathematical framework originally developed to describe quantum many-body physics, the behavior of huge numbers of interacting quantum particles. Multiverse's co-founder and chief scientific officer, Dr. Román Orús, pioneered applying that same math to the very different problem of squeezing down AI models. So the DNA here is genuinely quantum research, even though the product that just raised half a billion dollars runs entirely on classical computers. CEO Enrique Lizaso framed the bet like this in the company's announcement: quote, 'The AI industry has accepted a false constraint for years — that powerful models require expensive infrastructure. That constraint is gone. We have proven that AI can run at full performance on a smartphone, inside a sovereign data center, on a factory floor with no cloud connection. This round is the moment we scale that proof across every industry that needs it,' end quote. The money's earmarked for R&D, sovereign AI infrastructure — the idea that a country or company wants its AI running on its own hardware, under its own control, not routed through someone else's cloud — and expansion into manufacturing, finance, energy, aerospace, defense, and healthcare.

Now here's where it gets interesting, because the investors backing this round aren't just talking about a compression tool — they're describing Multiverse as basically the operating system for the AI industry's cost problem. Damien Henault, managing director at Forgepoint Capital International, argued that Multiverse has evolved into something bigger than a compression vendor. In the company's announcement, he said, quote, 'Multiverse sits at the intersection of the infrastructure and the application layers and has evolved from being the leading downstream LLM compression technology to becoming a complete AI foundry and Operating System. It is the only company we've seen that has both the technical foundation and the commercial traction to be that critical platform,' end quote. That's about as bold a claim as a venture investor makes — 'the only company we've seen' — and it tells you how much conviction is behind this check. BNP Paribas's Yann Lagalaye leaned into the sustainability angle, saying in the announcement, quote, 'This translates into lower GPU utilization, reduced energy consumption, and decreased infrastructure requirements, all while preserving comparable performance and accuracy levels,' end quote. And Bullhound's Per Roman framed it as a European champion story, quote, 'For years, Europe has worried it cannot compete at the frontier of AI. Multiverse is the answer — a team out of San Sebastián whose technology lets anyone run powerful models on their own terms, on their own soil, without depending on anyone else,' end quote. So is this a quantum computing win, or is it really just an AI efficiency story that happens to have quantum physics in its family tree? I think it's mostly the second one, and the skeptic's note attached to this story makes the same point plainly: this is fundamentally a quantum-inspired classical AI compression business, not a quantum hardware or quantum algorithms company. The huge round tells you a lot about investor appetite for AI efficiency plays that can wear the word 'quantum' — it doesn't tell you much about the state of actual quantum computers. There's also a numbers wrinkle worth flagging. Most outlets covering this — The Quantum Insider, Silicon Republic, FinSMEs — report that one-point-seven billion dollar pre-money figure. But at least one outlet, TechFundingNews, reports a two-point-three billion dollar post-money figure instead. Pick a number, any number — the likely explanation is these are two different ways of measuring the same round, pre-money versus post-money, but headline valuations already diverging across outlets before the round has even fully closed is exactly the kind of thing that should make you read valuation headlines with a raised eyebrow, on this story and every story like it. And here's the quiet part: none of this changes anything about the state of quantum hardware, error correction, or logical qubits. Multiverse's tensor-network math traces back to quantum many-body physics, but CompactifAI runs on ordinary GPUs and phones, today, with no quantum computer anywhere in the loop. So where does that leave us? Time for the Hype Check. I'm giving this one a five out of ten on substance. Here's my reasoning: the money is completely real, the investor list is serious, sovereign-grade institutional capital, and the underlying compression numbers — eighty to ninety-five percent model size reduction — are a genuinely useful engineering result if they hold up in production at scale. That's the substance. But the framing, that this is a 'quantum computing' story, is doing a lot of unearned work — this is an AI infrastructure company using math with quantum origins, full stop, and the size of the check says far more about how desperate the AI industry is to cut its GPU and energy bill than it says about where quantum computing itself stands. Strip the word 'quantum' out of the press release, and this is still one of the biggest AI infrastructure rounds of the year — which, honestly, might be the more interesting story.

These are threads we'll keep pulling on this week, especially as more 'quantum-inspired' classical companies chase this kind of capital, and as IonQ heads toward its own earnings test on August fifth. If the show's earning your time, hit follow wherever you're listening — it genuinely helps more people find this. 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!