Tech

AMD and PsiQuantum Expand Partnership to Build Quantum Computing Infrastructure

AMD

Quantum computing is moving from laboratory research towards the difficult business of building machines that can operate at meaningful scale. A new partnership between Advanced Micro Devices (AMD) and PsiQuantum, announced on 12 November 2026, highlights an important part of that transition: large quantum computers will still depend on powerful conventional computing systems to control, coordinate and support them.

AMD plans to make an equity investment in PsiQuantum as the companies expand their collaboration. They intend to co-design computing infrastructure for utility-scale, fault-tolerant quantum computers and develop applications that combine quantum processors with high-performance computing (HPC) and artificial intelligence. The companies have not disclosed the size of the planned investment, and the announcement does not mean a commercial, large-scale quantum computer is ready for customers.

What AMD and PsiQuantum announced

The expanded partnership brings together different parts of the computing stack. AMD is expected to contribute central processing units, graphics processing units, adaptive computing technology and software. PsiQuantum will contribute its silicon-photonics platform, which uses optical technology to move information between components.

The relationship builds on earlier work involving AMD Versal field-programmable gate array technology, which PsiQuantum selected for development of its fault-tolerant quantum-computing platform. The new arrangement is broader: rather than concentrating on one component, the companies say they will work across infrastructure and applications.

That distinction matters because a quantum computer is not simply a faster version of a desktop computer. It requires specialised quantum hardware, control electronics, conventional processors, software, cooling and a carefully coordinated flow of information. A useful system must make all of those elements work together reliably.

Close-up conceptual view of a photonic processor connected by optical light paths
Silicon photonics can use light to move data between computing components.

Why quantum computers need conventional computing

Quantum processors work with quantum bits, or qubits, whose behaviour is different from the binary bits used in ordinary computers. Their potential comes from the ability to represent and process certain kinds of information in ways that may offer advantages for carefully chosen problems. However, qubits are delicate, and useful calculations require complex control and error-management systems.

Conventional processors remain essential for preparing calculations, coordinating hardware, processing measurement results and running the software around a quantum workload. GPUs can also accelerate simulations and other computational tasks that help researchers develop and test quantum applications. The practical direction is therefore hybrid computing: classical and quantum resources working together rather than one immediately replacing the other.

For data-centre operators and research institutions, the challenge is not only obtaining a quantum processor. They also need networks, memory, storage, software tools and reliable ways to schedule work across different types of hardware. Partnerships that address these layers could be important to the commercial development of the field.

What silicon photonics contributes

PsiQuantum’s platform is based on photonics, a field that uses light to carry and process information. In silicon photonics, optical components can be integrated with semiconductor manufacturing techniques. The technology is relevant to quantum computing because scaling a complex system requires efficient ways to connect components and move information.

Optical links are already important in modern data centres, where moving information between processors can become a bottleneck. PsiQuantum’s approach applies photonic technology to its quantum-computing architecture, while AMD supplies a broad range of conventional compute and software capabilities. The partnership aims to bring those strengths into a more unified system.

It is important not to confuse a promising architecture with a finished product. Engineering, manufacturing, error correction, system integration and independent performance testing all remain necessary before customers can judge whether a particular design offers practical advantages.

Potential applications: science before everyday computing

Quantum computing research is often associated with problems involving complex molecules, materials, optimisation and cryptography. The AMD–PsiQuantum announcement points to applications that combine quantum workflows with HPC and AI. The companies said their existing work had produced initial results in simulations involving covalent inhibitors, a class of molecules relevant to oncology research.

That statement should be understood as an early research claim, not proof that a new medicine has been discovered or that a quantum computer has already outperformed established methods in clinical practice. Drug development involves many stages, including laboratory validation, safety testing and clinical trials. Computing can support research, but it does not remove those requirements.

In materials science, researchers hope quantum systems may eventually help model chemical and physical behaviour that is difficult to simulate accurately with conventional methods. Financial modelling and energy research are also frequently discussed as possible application areas. Each use case will need evidence showing that a quantum approach is more useful, reliable or cost-effective than available alternatives.

What the announcement does—and does not—tell us

The companies have confirmed an expanded strategic partnership and AMD’s intention to invest, but they have not published the investment amount or a firm delivery timetable for a utility-scale system. Readers should therefore treat the announcement as a step in infrastructure development, not as a launch date for a finished quantum computer.

The wider market is attracting attention from governments, universities and technology companies because quantum computing could eventually complement AI and conventional supercomputing. That interest is significant, but the field still faces difficult engineering questions. Fault-tolerant operation requires reliable qubits and error correction at scale, while commercial adoption depends on useful applications and manageable costs.

What businesses and technology leaders should watch

Organisations do not need to rush into buying quantum hardware. A more practical approach is to monitor developments in hybrid computing, identify research problems that may benefit from advanced simulation, and build a realistic understanding of data governance and specialist skills. Universities and research teams can also watch for opportunities to collaborate with hardware and software providers.

Useful milestones will include independently measured system performance, progress in error correction, software compatibility, access to developer tools and demonstrations that solve a clearly defined problem. The most convincing evidence will be repeatable results rather than broad claims about the number of qubits or theoretical computing power.

For readers following the wider technology landscape, Digismartiens has also covered the goals of the AI Campus, practical questions around AI in healthcare claims and tools for identifying synthetic media in its Google SynthID guide. These subjects show how new computing capabilities move from research into specific business and public applications.

The bigger picture

The AMD–PsiQuantum partnership reflects a broader shift in the quantum-computing industry: progress depends on the infrastructure around the quantum processor as much as on the processor itself. By combining AMD’s conventional computing portfolio with PsiQuantum’s photonic technology, the companies are trying to address some of the integration challenges involved in building larger systems.

Whether this collaboration produces a commercially useful advantage will depend on engineering progress and measurable results over time. For now, the announcement is best viewed as a meaningful infrastructure partnership in a field that remains promising, technically demanding and still under development.

Sources and further reading

Related Articles

Back to top button