Quantum computing reaches a historic turning point: bidding farewell to "physical experiments," supply chains and manufacturing become the key to success.

The quantum computing industry is undergoing a profound paradigm shift. Jefferies believes that the industry's core challenge has shifted from "whether we can build good qubits" to "whether we can use these qubits to build a truly usable computer." This shift means that the focus of competitive advantage is moving from pure physics innovation to a comprehensive shift in systems engineering, manufacturing control, and ecosystem development.

According to a report released by Jefferies' analyst team after attending the Quantum World Conference (QWC), the industry's "scoreboard" is being rewritten: logical qubit performance, error correction overhead, circuit depth, decoding latency, network interconnectivity, and system-level integration have become core metrics for measuring progress. Meanwhile, the U.S. Department of Energy's Office of Programs is leading the development of unified measurement standards, with the government's role evolving from a funder to an industry rule-maker.

The report also notes that the National Quantum Initiative Reauthorization Act (NQI Reauthorization Act) is expected to be passed soon, which will be a positive catalyst for companies that already hold federal contracts, including D-Wave Quantum (QBTS), Quantumuum (QNT), and Rigetti Computing (RGTI).

The implications of this shift for investors are clear and direct: companies that can build barriers in manufacturing repeatability, supply chain resilience, and system integration capabilities will gain a competitive edge in the commercialization race. Jefferies maintains a positive overall rating on the quantum sector, believing that the industry's transformation toward quantifiable, system-level milestones is providing investors with a clearer path to commercialization.

The government is reshaping industry standards, and federal funding is being linked to performance.

The quantum industry has long lacked a unified evaluation framework, but this situation is changing.

The Jefferies report points out that classical computers make an error approximately once every 10¹⁶ operations, while the target for complete fault tolerance in the quantum realm is currently defined on the order of 10⁶, with 10⁹ still considered a high level. The gap between "the existence of logical qubits" and "the superiority of logical qubits over physical qubits" remains the reality of the field.

The significance of government intervention lies in breaking the convention of corporate self-endorsement. Third-party projects now require companies to provide data such as logic error rate, gate fidelity, fault-tolerant operation, circuit depth, and effective scientific workload, making it difficult for companies to rely on self-selected benchmarks to get by. The funding model has also changed accordingly—upfront capital is used for tool development, and the remainder can only be unlocked upon reaching R&D milestones.

It is noteworthy that commercial validation is now driving government demand, rather than the previous government-led model. The executive order issued in June brought previously unassessed buyers of quantum technologies into the market. The National Quantum Initiative Reauthorization Act further requires the development of an international quantum cooperation strategy, includes the State Department on an interagency subcommittee, and extends the mandate to 2034. Jefferies believes that the supply chain provisions in the act directly support IonQ's (IONQ) strategy of pursuing vertical integration through SkyWater.

The rise of distributed architectures has brought quantum networks from concept to reality.

The problem of extending the boundaries of a single quantum processor has become one of the most discussed topics at this year's QWC.

Jefferies' report points out that as constraints such as control complexity, cooling requirements, cabling, laser transmission, calibration, yield, crosstalk, and physical size become increasingly apparent, more and more companies expect the final solution to involve multi-processor network interconnection—which is exactly the same logic as in classical computing where multiple processors form a cluster through a high-speed network.

This architectural shift has a significant impact on reshaping the value chain. Monolithic machines concentrate value in qubits, control systems, cryogenic refrigeration, and software; while distributed machines add entirely new infrastructure layers such as quantum storage, converters, photonic interconnects, switching, entanglement distribution, network control, and distributed compilation. Jefferies believes that IONQ, with its continued investment in network interconnects and photonic interconnects, is a major beneficiary of this trend.

Meanwhile, the heterogeneous architecture of CPU+GPU+QPU has been confirmed by QWC as the most likely final state. IONQ announced that its Superion 256 system will directly connect to NVIDIA GB200 NVL72 via NVQLink. The QNT technology stack also follows this logic, supporting the simultaneous execution of classical and quantum code, with the classical part executed on the GPU. This architecture embeds quantum computing into the existing high-performance computing and data center ecosystem, rather than requiring enterprises to start from scratch, thus significantly lowering the commercialization threshold.

Manufacturing and supply chain have become core competitive variables.

The focus of the competition in quantum computing is shifting from the laboratory to the factory floor.

Jefferies cited a global supply chain study covering 160 quantum computing companies, showing that 90% of these companies rely on at least one overseas supplier, and 74% serve overseas customers. As QPU production expands from 5 units to 50 or even thousands, remanufacturability, packaging processes, component availability, and laser stability will become key differentiating factors.

Supply chain bottlenecks are categorized into two types: first, quantum-specific components, which are scarce and some segments are essentially reliant on a single source; and second, standard electronic components, where the main issue is delivery time. Insufficient domestic production of key minerals such as indium and gallium in the United States, coupled with direct competition between quantum companies and hyperscale cloud service providers for commodity supplies for optical fibers, lenses, and micro-optical devices, highlights talent as the most challenging bottleneck in the supply chain.

In terms of response strategies, Jefferies believes that Quantinuum, IONQ, and Rigetti Computing have relative advantages. Quantinuum has established a reliable mass production path through multi-year supplier relationships and structured manufacturing collaborations without bearing the pressure of heavy asset self-built capacity; IONQ has achieved vertical integration through the acquisition of SkyWater, thereby controlling the technology roadmap and accelerating iteration speed; Rigetti Computing also owns its own wafer fabs, which helps control the roadmap and iteration pace.

Furthermore, Jefferies is optimistic about investment opportunities in the quantum "infrastructure" sector. Cryogenic refrigeration infrastructure is a key theme of this year's conference exhibition hall, and the report anticipates that advancements in cryogenic CMOS, high-density I/O, multiplexing, packaging, and automated testing will create considerable opportunities for traditional semiconductor and equipment suppliers.

Commercialization progress is evident at QWC, providing phased support for the industry's long-term narrative. Quantinuum announced the deployment of a Superion 256 system on the Florida International University (FIU) campus, effectively validating its localized system sales model. D-Wave Quantum (QBTS) announced a partnership with CGI, incorporating its Advantage2 system and hybrid solver into CGI's product portfolio. Jefferies points out that QBTS possesses an advantage that most gate model vendors lack—an off-the-shelf system that customers can use today for partially optimized workloads.

~~~~~~~~~~~~~~~~~~~~~~~

The above content is from Zhuifeng Trading Platform .

For more detailed analysis, including real-time updates and firsthand research, please join the [ Trading Channel Annual Membership ].

Risk warning and disclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.