Classiq exit: what Intel’s quantum deal teaches founders

On October 26, 2026, Intel signed a definitive agreement to buy Classiq, a Tel Aviv quantum software company founded about six years ago. Intel did not disclose the price. SiliconANGLE put it at $2.2 billion, and the Israeli business daily Calcalist reported $2.3 billion in cash and stock. Either way, the Classiq acquisition is one of the largest exits a quantum software startup has had, and it is Intel’s biggest deal since Mobileye.

Most coverage has focused on what Intel gets: a software layer to sit on top of its silicon spin-qubit hardware. For founders and university spin-outs, the more useful question is a different one. How does a company with roughly 150 employees and a little over $200 million raised end up worth more than $2 billion before quantum computers are commercially useful? Part of the answer is a platform that runs on everyone’s hardware. Another part is a patent portfolio that Classiq started building in its first two years. This article looks at both and at the lessons they hold for deep-tech founders.

The Classiq acquisition in brief

The basic facts, from Intel’s announcement and early press coverage:

• Buyer and target: Intel Corporation is acquiring Classiq Technologies Ltd., which builds software that turns high-level descriptions of quantum algorithms into optimized circuits.
• Price: not disclosed by Intel; reported at $2.2 billion by SiliconANGLE and $2.3 billion in cash and stock by Calcalist.
• Timing: expected to close in the first half of 2027, subject to regulatory approvals and customary conditions. Until then the two companies run separately.
• People: Intel CEO Lip-Bu Tan named co-founder and CEO Nir Minerbi, co-founder Yehuda Naveh and Amir Naveh in the announcement. Calcalist reported that employees hold about 10% of the company and that Intel set aside more than $100 million in retention grants.
• Customers: Classiq lists BMW, Rolls-Royce, HSBC, Deloitte and Citi among its users. Intel described “dozens of global enterprises.”

Intel said Classiq will keep supporting other vendors’ hardware after closing and will stay modality-agnostic. That promise matters, and we will come back to it.

Why a quantum software company commanded this price

Writing a quantum program today is closer to wiring logic gates by hand than to writing software. A developer has to decide which gates go where, how many auxiliary qubits to use, and how to fit the circuit to one machine’s connectivity and error rates. Classiq’s platform lets a developer describe what the algorithm should do in a high-level modeling language, and its synthesis engine produces a circuit optimized for a chosen machine. It also benchmarks hardware and estimates cost and performance before a job runs.

Intel compared this to electronic design automation (EDA) and computer-aided design in the chip industry. Nobody lays out billions of transistors by hand; synthesis tools do it. If quantum computing follows the same path, the company that owns the design layer gains leverage over every hardware vendor beneath it. SiliconANGLE made a similar comparison, likening the deal to Google’s 2005 purchase of Android.

Hardware independence is the core of that leverage. Classiq supports superconducting, trapped-ion, neutral-atom, photonic and spin-qubit machines, including those from IBM, Google and IonQ. Its investors reflect this: AMD Ventures, Qualcomm Ventures and IonQ all joined a strategic round in November 2025. A buyer acquiring a tool that every hardware maker relies on gets more than a product. It gets a seat at the center of the ecosystem.

The patent portfolio behind the platform

Classiq treated patents as a core asset early. In October 2021, about a year after its founding, it announced nine patent filings covering circuit synthesis, debugging circuits too large to simulate, and optimizing designs for particular hardware. It also had a director of IP strategy on staff, Shmuel Ur, an unusual hire for a startup that size.

Those filings matured. Public patent databases now list more than 20 U.S. patents and published applications assigned to Classiq Technologies, most of them issued in 2024 through 2026. The titles trace the platform’s architecture:

• “Functional-level processing component for quantum computers” and “Dynamic synthesis of gate-level implementations of functional blocks in quantum circuits,” covering the step from high-level functions to gates.
• “Quantum circuit synthesis using CSP” and “DAG-based CSP quantum circuit modeling,” covering synthesis treated as a constraint satisfaction problem.
• “Determining an implementation of a quantum program that has a minimized overall error rate” and “Selecting a quantum computer,” covering hardware-aware optimization.
• “Determining dynamic quantum error correction schemes” and “Auxiliary qubits analysis based on partial compilation of quantum programs,” covering resource and error management.

For an acquirer, a portfolio like this does two jobs. It makes it harder for a competitor to copy the synthesis approach directly, and it shows that the technology is novel enough to have survived examination at the U.S. Patent and Trademark Office.

Software claims that held up

Many founders assume software for quantum computers cannot be patented in the United States because of the limits on abstract ideas under 35 U.S.C. § 101. Classiq’s portfolio shows otherwise. The claims that tend to survive examination are tied to a concrete technical result: fewer qubits, a lower error rate, a circuit that fits a specific machine’s connectivity. Claims that recite only a mathematical method, without that technical improvement, are the ones that run into trouble. We cover the distinction in more detail in what you can protect with a quantum computing patent and in our article on quantum error correction patents.

Claims that stay valuable to any buyer

Classiq’s patents describe synthesis and optimization in terms that apply to any qubit technology. That broad, modality-agnostic framing is part of why the portfolio is worth a lot to Intel, which builds spin qubits, and would have been worth a lot to a buyer that builds trapped-ion or superconducting machines. A portfolio written around one hardware partner’s architecture narrows the list of acquirers. A portfolio written around the problem every vendor shares keeps that list long.

What diligence looks like in a deal like this

A deal of this size comes with intense IP diligence. The buyer’s counsel will check that the patents are valid and that they cover the product, but most of the work goes into ownership. Typical questions include:

• Chain of title: did every inventor, including founders, early contractors and university collaborators, sign an assignment to the company? Are those assignments recorded at the USPTO?
• Employee inventions: Israel’s Patents Law gives employee-inventors a possible claim to royalties for “service inventions,” heard by a royalties committee at the Israel Patent Office under section 134. Israeli companies usually address this in employment agreements with an express clause on royalties. An acquirer will read those agreements closely.
• Open-source and third-party code: which licenses apply to the code in the product, and do any of them restrict how the software can be combined with the buyer’s code or licensed to customers?
• Government and university rights: did any grant, government program or university lab fund the work in a way that leaves a license or other claim behind?
• Customer and partner agreements: do any contracts grant exclusive rights, assign improvements to customers, or end on a change of control?

Most of these items are cheap to get right at the start and expensive to fix during a deal. Our guide to patent due diligence for startups walks through the same list from the investor side. A freedom to operate analysis covers the other half of the picture: whether your product infringes someone else’s patents.

Lessons for startups and spin-outs

Classiq’s exit is unusual in size, but the IP habits behind it are not. Six of them carry over to almost any deep-tech company:

1. File before you present. Classiq’s first filings came within about a year of founding, before it was well known. In most countries a public talk, paper or demo can destroy patent rights unless an application is already on file. A well-prepared provisional application is often enough to secure the date; see our notes on provisional patent application mistakes and on when to file.
2. Make someone own IP strategy. A dedicated IP lead is rare in a 20-person startup, but someone on the team should own invention disclosures, filing decisions and a regular review of what competitors are patenting.
3. Claim the layer you own. Classiq doesn’t make hardware, and it didn’t try to patent hardware. Its claims focus on synthesis and optimization, the part of the stack it controls and every customer needs.
4. Write claims for more than one buyer. If your claims only make sense with one partner’s hardware, your exit options shrink to that partner and its competitors. Claim the general technical problem, with narrower fallback claims for specific implementations.
5. Keep the chain of title clean from day one. Signed invention-assignment agreements for every founder, employee and contractor cost little. Missing ones can delay a closing or reduce the price.
6. Treat the portfolio as a product. Patents granted in 2024 to 2026 came from applications filed years earlier. Continuations and new filings followed the product as it grew. A portfolio built this way tells a buyer the story of the technology.

What happens next

Intel expects to close in the first half of 2027, after regulatory review. Until then Classiq operates independently, and its integrations with other hardware vendors continue. Intel’s announcement lists the usual risks: approvals, keeping key staff and customers, and the uncertain timeline for commercial quantum computing.

For Israel’s quantum sector, the deal sets a new mark. Calcalist noted that the previous record for a quantum acquisition was IonQ’s purchase of Oxford Ionics for about $1 billion. A software exit at more than twice that price will draw investor attention to quantum software companies, which need less capital than hardware makers to reach scale.

For founders, the takeaway is simple. Classiq’s value came from technology that solves a problem every hardware vendor shares, and from a patent portfolio that described that technology broadly and early. Both started long before anyone was talking about an exit. If you are building in quantum computing or another deep-tech field and want to talk through your patent strategy,

References

1. Intel Corporation, “Intel to Acquire Classiq to Create a Full-stack Quantum Computing Platform,” October 26, 2026.
2. SiliconANGLE, “Intel buys quantum software pioneer Classiq for $2.2B,” October 26, 2026.
3. Quantum Computing Report, “Intel to Acquire Classiq: Strategic Move Creates Full-Stack Quantum Computing Platform,” October 2026.
4. JournalArta Global (summarizing Calcalist/CTech), “Intel’s $2.3 Billion Classiq Deal Would Set Quantum Record,” October 27, 2026.
5. The Quantum Insider, “Classiq Files Nine Patents for Quantum Algorithm Design,” October 19, 2021.
6. Justia Patents, patents and applications assigned to Classiq Technologies Ltd.
7. Quantum Computing Report, “Classiq Secures Strategic Funding from AMD, Qualcomm Ventures and IonQ,” November 14, 2025.
8. Israel Desks, “Israeli Quantum Software Company Classiq Secures USD 110m Boost,” May 2025.
9. Barnea Jaffa Lande, “Who Owns the Patent? Employees’ Right to Royalties,” February 2020.

This article gives general information, not legal advice. Deal terms are as reported at the time of writing and may change before closing.

This article is general information, not legal advice, and reading it does not create an attorney-client relationship. Patent law changes and every situation differs. Please talk to a patent attorney about yours.

Message Alexander on LinkedIn