On October 27th, Intel announced that it has reached a final agreement to acquire the Israeli quantum computing software startup Classiq Technologies. The transaction is expected to be completed in the first half of 2027, subject to regulatory approval and customary closing conditions.
Although the acquisition price has not been disclosed, according to informed sources, the transaction price is approximately $2.3 billion (approximately 15.3 billion yuan), and the payment method will be determined upon the completion of the transaction. If true, this will be the largest acquisition in the global quantum computing industry to date.
Intel CEO Chen Liwu stated in the statement that this merger will transform Intel's quantum computing division into a "full-stack quantum computing platform". Nir Minerbi, the co-founder and CEO of Classiq, and the entire team will officially join Intel.
This is the first acquisition announced by Chen Liwu as Intel's CEO in the past one and a half years, and it is Intel's second-largest acquisition in Israel, following the $1.53 billion acquisition of the Israeli autonomous driving technology company Mobileye in 2017.
"Hidden champion" in the quantum software field
Classiq was founded in May 2020, headquartered in Tel Aviv, Israel. It was jointly founded by Nir Minerbi, Yehuda, and Amir. The company currently has approximately 130 employees.
The core product of the company is a unified software platform that is independent of the underlying quantum bit mode and hardware. Through the Qmod quantum modeling language, deep compilation technology, and AI tools, it automatically combines advanced functional models into optimized quantum circuits. Developers do not need to deeply understand the underlying knowledge of quantum circuit design and can create scalable quantum applications.
This platform supports multiple cloud platforms such as AWS Braket, Microsoft Azure Quantum, and Google Cloud, and can directly integrate major quantum hardware such as IBM, IonQ, QuEra, and Quantinuum, and has over 60 core patents.
In terms of financing, Classiq has accumulated a total financing amount of more than $200 million. Among them, the $110 million C-round financing completed in May 2025 set a record for the largest single financing in the quantum software field, and in November of the same year, it received strategic investments of tens of millions of dollars from institutions such as AMD Ventures, Qualcomm Ventures, and IonQ.
In terms of customers, Classiq has established partnerships with BMW Group, Citibank, Deloitte, Rolls-Royce, Mizuho Bank, Toshiba, AT&T, and others. In 2023, the company collaborated with NVIDIA and Rolls-Royce to design a large-scale quantum computing fluid dynamics circuit for improving the efficiency of jet engines, with a measurement depth of over 10 million layers.
Building a full-stack capability of software and hardware collaboration
Intel has been investing in quantum computing research since 2015, with a core technology route based on silicon-based spin quantum bits. It is currently the only company that can use traditional silicon-based semiconductor manufacturing processes and equipment to mass-produce quantum bit chips and complex cryogenic control electronic devices.
On December 2024, Intel released the 12-qubit silicon spin quantum chip Tunnel Falls, manufactured using 300-millimeter wafers and EUV lithography technology, with a wafer-level yield of 95%. On July 2026, Intel announced a collaboration with Hitachi and the Japan Institute of Industrial Technology to develop a 100+ qubit silicon-based quantum chip based on Intel 18A process.
Classiq's synthesis engine and Qmod language complement Intel's strong hardware capabilities - Intel has compared it to the transformation of the traditional chip design industry by EDA software four decades ago, believing that it has driven the integration circuit from a niche profession to a large-scale industry through "abstracting hardware complexity".
After the transaction is completed, Intel's computing product portfolio will further extend from CPUs, GPUs, and ASICs to quantum processing units (QPU), forming a full-stack capability of hardware and software collaboration.
In the announcement, Intel promised that after the transaction, the Classiq platform will continue to maintain compatibility with multiple quantum hardware platforms, including those of competitors, to maintain its established open ecosystem positioning. This statement aims to alleviate customers and partners' concerns about the platform's "closedness".
Intel also emphasized that quantum computing is positioned as a technology that works in conjunction with traditional computing, rather than completely replacing the existing architecture. Chen Liwu previously stated that it is expected that quantum computing will have the opportunity to enter mainstream applications in the next three to five years, forming a heterogeneous hybrid computing network with CPUs and GPUs.
Chen Liwu's "Ten-Year Forward-Looking Project"
Chen Liwu included quantum computing in his "Ten-Year Forward-Looking Project" layout. He pointed out that the human brain processes extremely complex cognition with only a few watts of energy consumption. There is a "potential technical space of reducing power consumption by 10,000 times" for future computing power, and quantum computing is one of the key paths to break through the bottleneck of traditional computing power and energy consumption.
However, the acquisition of Classiq is only the latest link in Intel's AI computing power layout chain. "International Electronics Business Review" found through analysis that Intel's overall layout can be summarized as: "CPU returns to the core, dual-wheel drive of process and packaging, implementation of rack-level systems, expansion of physical AI and edge, binding customers with custom chips and foundries, and making quantum computing a long-term reserve".
1. Re-defining AI computing power: CPU is not a supporting role
Chen Liwu judges that AI is moving from training to inference, and then to intelligent agents and multi-agent workloads. The common CPU:GPU ratio in the past training stage was approximately 1:8, and is converging towards 1:4 or even 1:1. This means that the CPU is no longer just an accessory "for the GPU", but a key layer that undertakes task orchestration, tool invocation, data scheduling, and low-latency control.
2. Data Center: Intel Xeon 6+ targets Agentic AI
Intel released the Intel Xeon 6+ based on Intel 18A, with a maximum support of 288 energy-efficient cores, for cloud-native, Agentic AI, and network-intensive loads. Xeon 6 was also selected as the Host CPU of the NVIDIA DGX Rubin NVL8 system, indicating its position in the AI server ecosystem.
3. Process and Packaging: 18A mass production, 14A accelerated introduction
Intel 18A: Yield is improving month by month, has supported products such as Panther Lake, and received a large number of design orders.
Intel 14A: 0.5 PDK has been delivered, 0.9 PDK is expected to be released in October 2026, the risk pilot production schedule is between the second half of 2027 and the first half of 2028, mass production is in 2028-2029.
Advanced Packaging: EMIB/EMIB-T is positioned as the key capability对标ing CoWoS， and some projects have been evaluated by customers to be transferred to the EMIB platform.
4. Rack-level AI Infrastructure： From selling chips to selling systems
Intel， together with Foxconn and SambaNova， launched Rack Scale Blueprints， building a rack-level solution for AI inference.
"Vector Core Compute" adopts a decoupled inference architecture: the Intel Xeon CPU is responsible for orchestration and execution, SambaNova RDU is responsible for decoding, and the NVIDIA GPU is responsible for pre-filling. This indicates that Intel is no longer merely emphasizing a single chip, but is attempting to participate in the AI cabinet and system-level competition.
5. Physical AI and Edge: Pushing AI from Data Centers to Device Sides
Intel established the Client Computing and Physical AI business group to promote AI into physical systems such as robots, industrial automation, and edge devices. The product side includes the third-generation Core Ultra, Intel Robotics AI Suite, and the OpenVINO Physical AI software stack, emphasizing heterogeneous collaboration of CPU + GPU + NPU.
6. Custom Chips and Contract Manufacturing: Binding with Google, Tesla, AMD, etc. Customers
Promoting Intel to shift from "only making its own products" to "creating custom chips for customers". Relevant measures include long-term cooperation with Google to develop IPU; participating in the Terafab project under Musk's leadership to explore high-scale AI chip manufacturing; collaborating with AMD on x86 AI software optimization. The contract manufacturing customers also include Broadcom, Microsoft, and other potential or already signed customers mentioned in the market.
7. New Materials and Long-Term Technology: Glass Substrate, Diamond Heat Dissipation, Wide Bandgap Semiconductors
Actively laying out cutting-edge technology directions such as glass substrate, EMIB-T, gallium nitride, silicon carbide, indium phosphide, and artificial diamond heat dissipation to address the bottlenecks of high-performance computing chips in interconnection density, power consumption, and heat dissipation.
8. Quantum Computing: Completing the software layer to form a long-term computing power puzzle
After acquiring Classiq, Intel's quantum business will expand from hardware research and development to a hardware + software platform. Classiq's quantum modeling language and circuit synthesis capabilities help lower the threshold for quantum application development.
Disclaimer: The content is from the internet. In case of any dispute, please contact customer service.
