Why did the largest quantum acquisition go to a software company?

On October 26, 2026, Intel signed a definitive agreement to acquire Classiq, a quantum software company founded in Israel in 2020, for a reported ≈$2.3 billion (Quantum Computing Report, Oct 26, 2026; Ynet, Oct 26, 2026). If the figure holds, this will be the largest acquisition of a quantum computing company on record — about 2.1× the previous mark, IonQ's $1.075 billion purchase of Oxford Ionics in June 2025 (IonQ press release) — and the first time the record belongs to a company that owns zero qubits. Rosetta Quantum logs the dated facts, separates what is measured from what is announced, and will re-verify this entry when the deal closes or breaks.

What exactly was announced on October 26?

A signed agreement — not a closed acquisition. Intel's release describes a definitive agreement to buy Classiq, subject to regulatory approvals, with closing expected in the first half of 2027 (Intel, Oct 26, 2026). Until then the two companies will operate separately. The ≈$2.3 billion price, paid in a mix of cash and shares, comes from press reporting (Ynet/Calcalist, Oct 26, 2026), not from a figure printed in the release text we read. The fact sheet, source by source:

One governance detail, reported by Ynet citing sources close to the company: Intel CEO Lip-Bu Tan is a personal investor in Classiq and did not take part in the negotiations because of that interest.

Is this really the largest quantum acquisition to date?

Among disclosed prices, yes — by roughly 2.1×. The ladder of quantum M&A deals with a public number is short, and the reported Intel–Classiq price sits alone at the top. Note what the ladder can and cannot say: it only contains disclosed values, so undisclosed deals and mergers without a published price never enter it.

Why is the record check going to software, when the capital went to hardware?

Because the price is not buying qubits — it is buying tooling and customers. The capital map Rosetta published from QED-C data shows ≈87% of tracked private quantum investment going to hardware and only ≈6% to software (QED-C 2025, via our dated map). Against that backdrop, the largest check in the sector's M&A history just went to a company whose assets are a synthesis platform, a hardware-agnostic compiler layer and an enterprise customer list — not a single QPU.

The multiples say the same thing. Using Ynet's figures, the reported price will be more than 3× Classiq's last private valuation ($600–700M in 2025) and more than 23× its annual revenue (below $100M). Intel's own framing is sober: the deal moves its quantum group "from R&D toward a commercial business," and Intel's Pushkar Ranade cautioned that quantum will not be "a high-margin business from day one" (The Next Platform, Oct 28, 2026). That is a company paying a record premium for distribution in a market that, by its own admission, will not pay back quickly — in a sector where, as Rosetta measured in October, the last three Y Combinator batches produced zero quantum startups.

Where does Intel sit on the measured quantum map?

On the quiet end of it. Intel has invested in quantum since 2015 (Intel release), shipped the Horse Ridge cryogenic control chips and a quantum SDK starting in 2019, and built Tunnel Falls, a 12-qubit silicon spin-qubit QPU, with chip and control variants (18A-Q, 18A-C) of its 18A process in Oregon (The Next Platform, Oct 28, 2026). What Intel has not done is enter the independent funnels where hardware claims get measured: the six teams DARPA advanced to the final stage of its Quantum Benchmarking Initiative are Atom Computing, Diraq, IBM, IonQ, Microsoft and PsiQuantum (DARPA, Oct 7, 2026; our QBI explainer). The silicon-spin-qubit company in that list is Diraq — not Intel.

The promises are explicit and dated, and they belong in the future tense: Intel says the combination will create a full-stack quantum computing platform, will add QPUs alongside its CPUs, GPUs and ASICs, and will keep Classiq's multi-platform, modality-agnostic integrations (Intel release, Oct 26, 2026). All of that will exist, if at all, after a closing expected in the first half of 2027. Some of Classiq's named customers already run quantum programs on other vendors' machines — the on-premise buyer map Rosetta published in October shows who actually buys quantum computers today.

What will Rosetta track from here?

Three things, each falsifiable. First, the closing: Intel's release says first half of 2027; The Next Platform reports second half of 2027; the actual date, or the deal's termination, will be logged here. Second, the hardware-agnosticism vow: whether Classiq's platform still targets superconducting, trapped-ion, neutral-atom and photonic machines a year after closing, or narrows toward Intel silicon. Third, the capital map: whether quantum software's ≈6% share of tracked private capital (QED-C 2025) moves, or this record stays a one-off. Rosetta Quantum will re-verify this entry — same URL, updated verdict — when the deal closes, is amended or is terminated.

What we don't know

• The exact price. The $2.3B is press-reported (QCR; Ynet/Calcalist). The investor-relations copy of Intel's release we read on November 2, 2026 does not print a figure. If Intel's filings state a different number, this entry will be corrected.
• The closing date. Intel's release (via QCR) says first half of 2027; The Next Platform reports second half of 2027. We log the discrepancy rather than resolve it.
• Classiq's real revenue. "Below $100 million" (Ynet) is a ceiling, not a number; the >23× revenue multiple is therefore a floor.
• The deal's fine print. The cash/share split and the reported >$100M in retention grants (Ynet, citing sources) are not in any public filing we have read.
• Whether it closes. Regulatory approval is pending. Until closing, the measured fact is a signed agreement, not an acquisition.
• What it proves about the sector. One record deal is one data point; it does not, by itself, re-rate quantum software.

Rosetta Quantum publishes verdicts with raw, dated, reproducible data. This is educational content — not a product claim, and not investment advice.