The arrival of the QPU era may be faster than anyone expects.

In a recent interview, former Intel CEO Pat Gelsinger stated that quantum computing will become widespread within two years, accelerating the burst of the AI bubble, and will completely replace GPUs before 2030. In his view, quantum computing will form a 'Holy Trinity' of future computing alongside classical and AI computing.

On December 9, quantum hardware startup QuantWare officially unveiled its new quantum processor (QPU) expansion architecture, VIO-40K, with plans to achieve large-scale quantum chip production by 2026. At its Delft headquarters in the Netherlands, the company is actively building Kilofab—the world's largest and first dedicated wafer fab for quantum chip production. This is expected to boost production capacity by 20 times compared to current levels.

In early 2025, NVIDIA CEO Jensen Huang joked that 'quantum technology is at least 20 years away from practical applications.' However, by the end of the year, quantum startups and tech giants had unveiled increasingly aggressive roadmaps, with media and institutional forecasts turning optimistic.

Undoubtedly, 2026 will be a pivotal year for quantum computing to move toward practical applications. Now is the perfect time to review the technology and market landscape.

Let's first examine what happened in the quantum industry in 2025.

01

Quantum Computing Explodes in 2025

The key players in the quantum computing arena in 2025 can be summarized as the Big Three, the Quantum Four Heroes, and NVIDIA.

Google: Following the launch of its 105-qubit Willow superconducting quantum processor in late 2024, Google officially announced the achievement of 'verifiable quantum advantage' in October 2025. By running the 'Quantum Echo' algorithm, the Willow chip processed disordered time-correlated tasks 13,000 times faster than classical supercomputers. Experimental data further showed that as the number of physical qubits increased, the system's logical error rate decreased exponentially, validating the effectiveness of error correction theory at the physical level.

Ecosystem-wise, Google deepened its collaboration with NVIDIA, leveraging the CUDA-Q platform for large-scale physical simulations to address noise design challenges in next-generation processors. Additionally, Google partnered with the UK's National Quantum Computing Centre (NQCC) to provide cloud access to the Willow chip for British research institutions, supporting algorithm testing in fields like materials science.

IBM: In 2025, IBM continued its hardware iteration roadmap, focusing on enhancing processor performance and reducing control system costs. Hardware-wise, IBM released and delivered the 120-qubit 'Nighthawk' processor, featuring next-generation tunable couplers that improved computational performance by about 20% compared to its predecessor, 'Heron.' Simultaneously, IBM introduced the experimental 'Loon' chip to validate the stability of large-scale fault-tolerant components.

In terms of engineering control architecture, IBM announced its collaboration with AMD in October of the same year. Utilizing AMD's existing commercial FPGA chips, the partnership achieved real-time error correction control for qubits. Tests demonstrated that the solution met real-time error correction demands, with the project completed a year ahead of schedule. This progress proved that general-purpose commercial chips could replace custom hardware for quantum control, helping reduce the engineering costs of building fault-tolerant quantum computers and supporting IBM's 2029 fault-tolerant model plans.

Microsoft: In February 2025, Microsoft unveiled its first quantum chip, 'Majorana 1,' based on topological superconductor materials. This marked Microsoft's long-term investment in topological quantum computing transitioning from theoretical research to hardware prototyping. Microsoft stated that the chip leveraged new material properties to achieve immunity to environmental noise at the physical level. Despite its small qubit scale, the prototype validated the feasibility of topological protection mechanisms at the hardware level, providing an experimental foundation for future expansion.

Quantum Four Heroes (IonQ, Rigetti, D-Wave, QCI): In 2025, publicly traded quantum computing companies like IonQ, Rigetti, D-Wave, and Quantum Computing Inc. (QCI) underwent market consolidation and business adjustments. IonQ completed its acquisition of UK startup Oxford Ionics in June for approximately $1.1 billion, aiming to integrate ion trap technology patents and expand its engineering team. Rigetti Computing continued to advance the modular deployment of superconducting systems, optimizing the interconnectivity performance of its Ankaa-3 processor. D-Wave highlighted the hybrid solving capabilities of quantum annealing technology in logistics scheduling and supply chain optimization. Meanwhile, Quantum Computing Inc. continued its exploration in photonics, striving to lower the environmental thresholds for system operation.

NVIDIA: In 2025, NVIDIA underwent a strategic correction in the quantum field. Jensen Huang shifted from questioning quantum computing at the beginning of the year to publicly apologizing and establishing its strategic importance at the GTC Summit in March, followed by rapid capital operations.

In September, NVIDIA's NVentures heavily invested in three technical routes within a week: participating in Quantinuum's nearly $600 million financing round (ion trap), backing QuEra (neutral atoms), and co-investing in PsiQuantum's $1 billion Series E round (photonic quantum). This multifaceted approach aimed to cover mainstream hardware modalities through capital rather than developing its own QPU.

At the year-end GTC Conference in Washington, NVIDIA introduced NVQLink, enabling direct communication between quantum computer QPUs and GPUs. Huang pointed out that QPUs for quantum computing had received unprecedented support, with 17 quantum computing companies and eight US Department of Energy (DOE) national laboratories now integrated into NVIDIA's ecosystem.

02

2026: A Year Full of Promises

Based on the technical roadmaps of major vendors and predictions from industry analysts, 2026 is seen as a critical turning point for quantum computing to transition from engineering validation to utility verification. The industry's focus will shift from merely expanding physical qubit scales to validating logical qubit quality and deploying hybrid computing architectures in practical settings.

In the superconducting route, IBM has set 2026 as a key year to demonstrate 'quantum advantage.' The company plans to use its 'quantum-centric supercomputing' architecture to prove cost or precision advantages over classical computers in specific scientific tasks. Its processors are expected to support deeper quantum gate operations (targeting around 7,500 gates) to run more complex algorithms. Google, meanwhile, faces the engineering challenge of transitioning from error correction principle validation to building long-lived logical qubits, with a focus on further improving physical qubit coherence times and gate fidelity.

In emerging routes, neutral atom vendor QuEra Computing has set a goal to release a system with 100 logical qubits; photonic quantum vendor PsiQuantum is accelerating its ultra-large-scale system assembly in Chicago and Brisbane, expecting to enter a critical system integration phase in 2026; and QuantWare's plan to launch the Kilofab wafer fab aims to advance the industrial-scale production of quantum chips. All these plans carry high engineering uncertainty, making 2026 a test period to verify whether these ambitious goals can be achieved.

With the promotion of middleware platforms like NVIDIA's CUDA-Q, 'quantum-classical hybrid computing' is expected to become a standard deployment model in data centers by 2026. QPUs will be increasingly integrated as accelerators into high-performance computing clusters to handle specific simulation or optimization tasks. The widespread adoption of this architecture will drive the integration of quantum computing with AI workflows, particularly in large-scale model training optimization and complex molecular simulation, with the industry anticipating more application tests based on hybrid computing power.

In the cloud services sector, IBM, AWS, and Microsoft currently offer quantum access services. By 2026, cloud providers may further integrate 'quantum + classical' hybrid resources. The service model will evolve from providing access to single experimental hardware to offering hybrid computing environments integrated with high-performance computing resources. The threshold for enterprise users to leverage quantum computing power via cloud platforms to solve practical problems is expected to lower, but whether this can generate positive returns on investment in large-scale commercial scenarios remains to be seen.

At the security level, the US government is mandating the upgrade of digital infrastructure to be quantum-resistant. Under the strategic deployment of the National Security Memorandum (NSM-10), US federal agencies have entered the substantive implementation phase of migrating to post-quantum cryptography (PQC). The White House requires agencies to accelerate the phase-out of traditional encryption algorithms like RSA and has set 2025-2030 as the critical window for core systems to complete PQC upgrades. This initiative aims to establish US dominance in formulating next-generation encryption standards and drive the reconstruction of defense systems in key sectors like finance and defense.

This also reflects the fact that the quantum computing competition in 2026 will not only occur at the enterprise level but also among major global nations.

03

Quantum Industrial Systems: A Global Race

Looking ahead to 2026, as major global economies enter a new policy cycle, the development logic of quantum computing is undergoing profound changes. Countries are shifting their focus from early competitions based on single scientific research metrics to building resilient industrial systems and mastering core links in the industrial chain.

In 2026, the United States will shift its strategic focus to strengthening its internal industrial foundation. Addressing the workforce gap risk highlighted by the nonprofit 'Quantum Leap,' the National Institute of Standards and Technology (NIST) is expected to accelerate the cultivation of quantum industrial clusters in Colorado and other locations. Unlike previous funding primarily for basic research, the fiscal year 2027 R&D budget indicates a policy shift toward nurturing skilled technicians and engineers and exploring new support models, including equity-for-funding, aiming to build a complete domestic industrial closed loop (closed loop) from R&D to manufacturing to address talent shortages in cryogenic electronics and microwave engineering.

2026 marks the beginning of China's 15th Five-Year Plan, with quantum technology positioned as a 'new growth engine' in relevant planning recommendations, suggesting that China's quantum industry is accelerating its transition from laboratory validation to industrial cultivation. At the infrastructure level, China's industry is exploring the integration of quantum computing into the national-level computing network, leveraging the 'East Data, West Computing' project. Companies like CT Quantum and Origin Quantum are expected to further pilot 'four computing integration'—quantum + supercomputing + intelligent computing + general-purpose computing—attempting to use real demands from public scenarios like weather forecasting and power grid scheduling to drive technological iteration.

In terms of industrial chain construction, Chinese enterprises are committed to building a fully autonomous ecosystem. Vendors like Quantum Xi (a representative firm) are advancing the R&D and validation of key components such as dilution refrigerators and special cables. Meanwhile, the photonic quantum route is also accelerating its layout. Companies like Turing Quantum are exploring differentiated applications in AI and biomedicine, while Tengjing Technology provides a solid foundation for this route through the R&D of precision optical components.

Europe and the UK are also accelerating their 'technological sovereignty' strategies. The UK has pledged sustained funding to support its quantum industry, relying on local companies like ORCA Computing to build internationally competitive industrial hubs. The EU, on the other hand, tends to support local companies in constructing quantum chip foundry facilities, attempting to gain more discourse power (voice) in chip manufacturing and establish an industrial 'third pole' with independent innovation capabilities.