AMD and PsiQuantum Expand Hybrid AI-Quantum Partnership

NewsTree24 | November 14, 2026 — AMD and quantum-computing company PsiQuantum have expanded their strategic partnership to design computing infrastructure for future utility-scale, fault-tolerant quantum systems. The November 12 announcement connects conventional high-performance computing, AI accelerators and photonic quantum technology, with drug discovery among the initial application areas the companies plan to explore.

What the expanded partnership includes

The companies said their work will build on PsiQuantum’s existing use of AMD Versal field-programmable gate arrays (FPGAs). The expanded relationship adds AMD Instinct GPUs, EPYC CPUs and the ROCm software stack to PsiQuantum’s silicon-photonics platform.

AMD also said it plans to make an equity investment in PsiQuantum. Neither company disclosed the investment amount in the announcement, and no commercial launch date for a complete utility-scale quantum computer was provided.

Why combine AI, high-performance and quantum computing?

Quantum computers will not simply replace conventional servers or GPUs. Large-scale systems still need classical computing for control, simulation, error correction, data processing and the surrounding software workflows. AI and high-performance computing can also help researchers model scientific problems and prepare algorithms for future quantum hardware.

PsiQuantum and AMD said their initial collaboration will focus on computational chemistry and drug discovery. The companies reported early results from using AMD GPUs to simulate covalent inhibitors, a class of molecules relevant to oncology research. That is a company-reported research result, not evidence that a fault-tolerant quantum computer has already delivered a new medicine.

What is technically significant?

The announcement describes work on a machine-learning decoder called PRISM, which the researchers have optimised for low-latency implementation on programmable hardware accelerators. In a photonic quantum computer, error correction and control are critical engineering challenges. Faster classical processing could help manage those tasks as a quantum system grows, although the practical performance of a complete system still has to be demonstrated.

The partnership also aligns with the US Department of Energy’s Quantum Genesis programme, which identifies scientific applications — including chemistry and materials research — that could help define what useful fault-tolerant quantum computing must achieve.

What enterprises and researchers should watch

  • Demonstrated workloads: look for reproducible benchmarks on chemistry, materials science and other target applications.
  • Hybrid software: assess how classical HPC, AI and quantum workflows will be orchestrated and supported.
  • Error correction: track evidence that control and decoding methods can scale beyond laboratory demonstrations.
  • Commercial milestones: distinguish a strategic partnership and planned investment from a deployed, commercially available quantum system.

For now, this is an infrastructure and research collaboration. Its significance lies in bringing AMD’s established compute platforms into PsiQuantum’s long-term effort to build fault-tolerant photonic quantum computers.

Related reading: the SGLang consortium’s work on open AI inference and more AI and technology coverage on NewsTree24.

Sources

Featured image is illustrative optical-laser equipment, not PsiQuantum hardware. Photo: HC Optical via Unsplash.

Leave a Comment