Intel agreed Monday to acquire Classiq Technologies, a Tel Aviv-based quantum computing software company, for approximately $2.3 billion in cash and Intel stock — the largest acquisition ever recorded in the global quantum computing industry, breaking by more than double the previous record set by IonQ’s roughly $1.075 billion purchase of Oxford Ionics. The deal, announced one day before Intel’s third-quarter earnings report, signals that the chip maker intends to own the software abstraction layer above a hardware race that remains unsettled — a strategic bet shaped by CEO Lip-Bu Tan’s career-long belief that software tools, not processors alone, determine who wins a computing platform.

Intel and the Cadence Playbook

Lip-Bu Tan did not come to Intel from a hardware company. He led Cadence Design Systems from 2009 to 2021, transforming it into the dominant supplier of electronic design automation (EDA) software — the tools that allow chip engineers to specify a circuit’s function at a high level while the software generates the physical gate-level implementation. Cadence’s revenue more than doubled under Tan. He earned a B.S. in physics from Nanyang Technological University and an M.S. in nuclear engineering from MIT.

When Tan describes Classiq, he uses the Cadence analogy explicitly. He told a September 2026 podcast that Classiq is “trying to be the next Cadence to computers, quantum computing.” His thesis: quantum computing will eventually need an abstraction layer above its hardware, just as classical chip design needed Cadence, and that layer will be as structurally valuable — and as sticky — as EDA became in the semiconductor industry.

In the early 1980s, designing a chip meant specifying every transistor interaction by hand. By the late 1980s, EDA software had made that infeasible task tractable, and the companies that controlled EDA tools — Cadence, Synopsys, Mentor Graphics — became indispensable infrastructure that chip designers could not build without. Tan spent twelve years running the dominant one. Now he is applying the same logic to a field where the hardware is not yet commoditized, the error correction is not yet reliable, and the software layer has not yet been claimed.

Classiq will become Intel’s global quantum development center after the deal closes, expected in the first half of 2027, pending customary regulatory approvals.

What Classiq Actually Does: Inside the Synthesis Engine

Classiq does not build quantum processors. What it has built is a three-layer quantum software stack designed to solve the field’s most immediate practical bottleneck: programming those processors.

At the top layer sits Qmod, Classiq’s native high-level quantum modeling language, accessible through a Python SDK and a web-based integrated development environment (IDE). A developer using Qmod specifies what computation a quantum algorithm should perform — using quantum functions, quantum objects, and high-level logic — without specifying which physical gates to use or how to wire them. This is analogous to writing Python or C++: the programmer expresses intent, not machine instructions.

Below Qmod sits the synthesis engine, Classiq’s core proprietary technology backed by more than 60 filed patents. The synthesis engine receives the Qmod model together with hardware constraints — maximum qubit count (circuit width), maximum gate layers (circuit depth), and target backend — and runs a combinatorial search across the exponentially large space of possible quantum circuit implementations. It outputs the optimal gate-level circuit for the specified constraints, in standard QASM format (Quantum Assembly Language), the lingua franca that quantum hardware backends can accept. Each backend has a different native gate set, qubit connectivity map, coherence time, and error profile. The synthesis engine transpiles separately for each — the same algorithm compiles differently for IBM’s superconducting backend than for IonQ’s trapped-ion machines.

That output then flows to the execution layer, where Classiq’s platform sends compiled circuits to quantum hardware via cloud APIs: AWS Braket, Microsoft Azure Quantum, Google Cloud, IBM Quantum, IonQ, and other backends. This architecture matters more as quantum hardware scales. With today’s hundreds of qubits, manual circuit design is already difficult. With tomorrow’s thousands or millions of qubits, the search space of possible circuit implementations grows exponentially — no team of human programmers can navigate it by hand. Classiq’s automated synthesis is designed to be the compulsory layer that makes those larger systems programmable at enterprise scale.

Enterprise customers including BMW, Citi, Deloitte, Rolls-Royce, Mizuho, HSBC, and Toshiba currently use the platform to design and test quantum algorithms for applications in financial modeling, logistics optimization, drug discovery, and materials science.

Why Silicon Qubits and Why Intel Believes It Has a Manufacturing Moat

Intel’s own quantum hardware effort takes a different technical path from the industry’s dominant approaches. IBM and Google have built their quantum systems around superconducting qubits — circuits cooled to near absolute zero that use the quantum properties of superconducting materials to encode information. Those systems have achieved the most documented progress and the largest cloud user bases.

Intel is pursuing silicon spin qubits: quantum bits encoded in the spin state of a single electron (spin-up = |1⟩, spin-down = |0⟩) confined in a silicon quantum dot. The physical attraction is manufacturing compatibility. Silicon is the substrate of every transistor Intel has ever made. Its Tunnel Falls research chip — a 12-qubit device — was fabricated at Intel’s D1 transistor facility in Hillsboro, Oregon using extreme ultraviolet (EUV) lithography and standard CMOS-compatible techniques, the same manufacturing processes the company uses to make conventional processors.

Silicon spin qubits are orders of magnitude smaller than superconducting qubits. Superconducting circuits typically occupy millimeter-scale footprints; silicon spin qubits measure approximately 50 nanometers by 50 nanometers — roughly one million times smaller. That compactness means that, in principle, millions of qubits could be fabricated on a single silicon wafer using existing foundry infrastructure — an economic and scaling proposition that superconducting approaches cannot easily match.

The challenges are real and unsolved. Silicon spin qubits face “valley splitting” problems in silicon-germanium substrates, coherence times that must improve significantly before deep quantum circuits become practical, and wiring density constraints that require cryogenic control electronics. Analysts noted in May 2026 that Intel’s quantum roadmap remains “still opaque,” and that the company must “scale funding and talent to catch up with rivals that have already delivered usable quantum cloud platforms.”

In July 2026, Hitachi launched a government-backed collaboration with Intel’s Japanese subsidiary and Japan’s National Institute of Advanced Industrial Science and Technology (AIST), funded by NEDO, Japan’s New Energy and Industrial Technology Development Organization. The project targets a 100-qubit silicon quantum computer implementing quantum error correction by fiscal year 2028, rising to a 1,000-qubit system by fiscal year 2030, using Intel’s 18A manufacturing process. In February 2025, Intel and AIST had already signed a memorandum of understanding targeting tens of thousands of qubits by the early 2030s.

By acquiring Classiq, Intel pairs this hardware roadmap with a software capability that already compiles for all competing hardware modalities. Whatever machine reaches fault-tolerant scale first — Intel’s silicon, IBM’s superconducting, IonQ’s trapped-ion, or a platform not yet dominant — a Classiq synthesis engine under Intel’s ownership can, in principle, compile for it.

After the Announcement: An Open-Ecosystem Pledge Worth Watching

Intel has committed to maintaining Classiq’s hardware-agnostic, multi-platform integrations after the acquisition closes. This pledge is strategically central: the value of Classiq to enterprise customers — BMW, Citi, Deloitte — rests entirely on its ability to target the best available quantum hardware rather than only Intel’s own. If Intel progressively restricted Classiq’s backends to favor its own silicon, those customers would have a powerful incentive to migrate to alternative platforms.

Classiq itself flagged this tension in a 2022 blog post, noting that the top consideration enterprise customers cited when choosing a quantum vendor was “forward compatibility, a hardware-agnostic, interoperable approach” — and that vertical integration creates risk when customers cannot freely switch hardware.

The pledge also exists against a backdrop that raises legitimate scrutiny. In 2025, Intel’s Data Center Group Vice President Kevork Kechichian announced a restructuring of Intel’s open-source software contribution strategy, explicitly stating that Intel would no longer provide contributions that give Intel’s competitors an advantage. Hyperscan, previously an open-source Intel project, transitioned to a licensed platform. This shift has drawn comparisons to Nvidia’s proprietary CUDA ecosystem, which created deep hardware lock-in for AI developers. Intel’s Classiq open-ecosystem commitment moves in the opposite direction. Enterprises currently using Classiq’s platform are well-positioned to monitor whether Intel’s stated policy aligns with Classiq’s software licensing and backend-access decisions over the coming years, particularly as Intel’s own silicon approaches practical qubit counts.

The Investor Windfall and What It Signals

The acquisition delivers decisive returns to Classiq’s six-year investor base. Entrée Capital, which led the company’s pre-seed round and remained its largest shareholder through the $110 million Series C in May 2025, stands to receive more than $300 million on an initial investment of just a few million dollars. Team8 Capital holds approximately 9% of Classiq, a stake worth roughly $207 million at the acquisition valuation. The Phoenix, the Israeli insurance and financial services group, holds a 7.5% stake valued at approximately $172.5 million — meaning Israeli retail savers with assets in The Phoenix’s funds will benefit from the exit.

Classiq’s roughly 150 employees collectively hold approximately 10% of the company, a stake worth about $230 million at the acquisition price. Intel has allocated more than $100 million in additional retention grants to keep key technical staff in place post-close. Each of the three co-founders — CEO Nir Minerbi, CPO Amir Naveh, and CTO Prof. Yehuda Naveh — is expected to receive between $80 million and $100 million from the transaction.

The co-founders’ backgrounds illuminate the company’s technical foundation. Minerbi and Amir Naveh are graduates of the Israel Defense Forces’ Talpiot program, the IDF’s elite scientific research unit that produces disproportionate numbers of Israel’s deep-tech founders. Prof. Yehuda Naveh — Amir’s father — spent approximately two decades at IBM’s Research Center in Haifa, where he founded and led IBM’s quantum computing software team as part of the global group that built IBM Q and its software environment. IBM, the institution from which Classiq’s CTO built his career, is also Classiq’s primary commercial competitor; estimates cited by industry sources suggest Classiq may have already surpassed IBM in quantum software market share, depending on how that market is measured.

Because a significant portion of the consideration is in Intel stock rather than cash, the precise proceeds for each stakeholder will depend on Intel’s share price at closing. Intel’s stock has risen substantially year-to-date in 2026, one indicator of CEO Tan’s Intel turnaround.

What Is the Quantum Computing Industry’s State in Late 2026?

For context: as of this writing, approximately 117 quantum computers exist worldwide. They remain expensive, largely inaccessible, and not yet commercially viable for most enterprise applications. According to the OECD’s March 2026 report, the field is “highly promising and highly uncertain.” Forrester’s 2026 analysis designates this year as the “fault-tolerant foundation era” — architectural progress toward error-corrected computation is real, but Forrester estimates fully fault-tolerant quantum computers capable of transformational commercial value remain at least five years away.

GQI (Global Quantum Intelligence), the research firm that publishes the Quantum Computing Report, has forecast that the mid-stack quantum software segment will expand at roughly 50% compound annual growth, exceeding $2 billion in annual software spend by 2030. Intel, through Classiq, would enter that market as its largest single entity.

McKinsey has separately estimated quantum computing could generate up to $2.7 trillion in economic value by 2035.

The quantum computing hardware market is divided among superconducting circuits (IBM, Google), trapped ions (IonQ, Quantinuum), photonics, neutral atoms, and silicon spin qubits (Intel). No approach has emerged as a clear scalability winner. Error correction and control systems sufficient for fault-tolerant computation remain active research problems for all modalities. In that environment, a platform that compiles for all of them is — as Tan frames it — a bet on infrastructure above the hardware race, not a bet on who wins the race.

BCG’s 2026 enterprise adoption survey found that more than 60% of surveyed enterprises experimenting with quantum computing had moved beyond exploratory projects to application-scale use cases. Survey firm Omdia found that 39% of quantum computing vendors expect their customers to begin production use of quantum computers in 2026, up from 14% in 2025.

Tan himself acknowledged the uncertainty with characteristic candor in his September podcast interview: “The verdict is still out. It’s still very early.”

Frequently Asked Questions

What does Classiq’s platform actually do, and why does Intel need it?

Classiq has built a software layer — comprising a high-level quantum programming language (Qmod), an automated synthesis engine, and hardware integration APIs — that compiles quantum algorithms for any quantum hardware backend (IBM, IonQ, Google, Intel’s own chips, and others) without requiring users to manually design quantum circuits at the gate level. Intel needs it because its own silicon spin qubit hardware, while promising for large-scale manufacturing, is still several years from commercially viable qubit counts. Classiq gives Intel a software position it can deploy immediately across the existing quantum ecosystem — and eventually link to Intel’s own silicon as it matures. Without this layer, enterprises cannot realistically program quantum computers as hardware scales.

What is the “Cadence of quantum” thesis and why does it matter to enterprise IT teams?

Cadence Design Systems became indispensable infrastructure in classical chip design by owning the EDA (electronic design automation) software layer — tools that translate high-level circuit specifications into manufacturable silicon. Intel CEO Lip-Bu Tan, who ran Cadence from 2009 to 2021, believes quantum computing will develop the same dependence on an abstraction-layer software platform. If he is right, the enterprise IT teams that commit to Classiq’s platform now may find themselves structurally dependent on Intel-owned quantum software the way chip designers became dependent on Cadence — with switching costs that rise as the software becomes deeply integrated into research workflows.

Will Classiq still work with IBM, IonQ, and Google hardware after Intel acquires it?

Intel has publicly committed to maintaining Classiq’s hardware-agnostic, multi-platform integrations after the deal closes. However, enterprise customers should note that this pledge exists alongside Intel’s 2025 shift away from its historic open-source software philosophy — Intel’s VP announced the company would restructure contributions to no longer benefit competitors. Classiq itself previously noted that “forward compatibility” and “hardware-agnostic interoperability” are the top considerations enterprise customers cite when choosing a quantum vendor. Whether Intel’s open-ecosystem pledge holds as its own silicon spin qubit technology matures is the central long-term question for the platform’s enterprise user base.

When will this deal close and what changes immediately for Classiq customers?

The acquisition is expected to close in the first half of 2027, subject to regulatory approval. Until then, Intel and Classiq operate as independent entities, and Classiq’s platform continues under its current terms and integrations. No changes to customer contracts or backend access are anticipated before close.