PsiQuantum and AMD Expand Partnership on Quantum Computing Infrastructure

Expanded partnership between PsiQuantum and AMD

PsiQuantum and Advanced Micro Devices (AMD) announced an expanded strategic partnership aimed at building the classical‑computing backbone for utility‑scale, fault‑tolerant quantum systems. The collaboration, first announced earlier this year around AMD’s Versal FPGA technology, now encompasses AMD’s Instinct GPUs, EPYC CPUs and the ROCm software stack, alongside PsiQuantum’s silicon‑photonic quantum platform. Both companies also disclosed that AMD will make an equity investment in PsiQuantum, underscoring a deeper financial commitment.

Co‑designing a hybrid computing fabric

The joint effort targets a seamless integration of three computing domains:

• Quantum photonics – PsiQuantum’s proprietary silicon‑photonic qubits, designed for large‑scale error‑corrected operation.
• High‑performance classical hardware – AMD’s GPUs for AI‑driven workloads, CPUs for orchestration, and FPGAs for low‑latency control.
• Software stack – AMD’s ROCm ecosystem will be adapted to support quantum‑classical workflows, including real‑time error‑correction decoders.

By aligning these layers, the partners hope to create a “hybrid fabric” where quantum algorithms, classical simulations and AI models can exchange data with minimal latency. The first application focus will be pharmaceutical research, leveraging existing GPU‑based simulations of covalent inhibitors and a machine‑learning decoder (named PRISM) that can run on programmable accelerators for real‑time error correction.

Early technical highlights

Although the announcement is primarily a partnership update, it references concrete technical progress:

• GPU‑accelerated chemistry – PsiQuantum’s computational‑chemistry workflows have already outperformed state‑of‑the‑art calculations on AMD hardware, delivering promising results for drug‑discovery targets.
• Machine‑learning decoder – The PRISM decoder, trained on photonic error patterns, has been optimized for low‑latency deployment on AMD’s programmable accelerators, a step toward the real‑time error correction required for fault‑tolerant operation.

These achievements illustrate how classical AI and HPC resources can amplify quantum‑ready algorithms even before a full fault‑tolerant machine is built.

Broader ecosystem and next steps

The partnership aligns with the U.S. Department of Energy’s Quantum Genesis priority applications, which highlight fault‑tolerant quantum computers as essential for solving problems in materials, energy, finance and health. Both firms plan to showcase further details at the SC26 conference in Chicago (Nov 15‑20) and are exploring a joint presence at the Illinois Quantum and Microelectronics Park (IQMP), where PsiQuantum is an anchor tenant.

In the longer term, the collaboration could set a template for other quantum‑hardware startups seeking classical‑computing partners. By combining photonic qubits with industry‑standard CPUs, GPUs and software, the ecosystem moves closer to the integrated stacks needed for practical quantum advantage.

---

Takeaway: The expanded PsiQuantum‑AMD partnership broadens the classical‑hardware and software toolbox supporting future fault‑tolerant photonic quantum computers, with an immediate focus on drug‑discovery simulations and real‑time error‑correction decoders. While no new quantum hardware metrics are disclosed, the joint effort signals a maturing approach to building the hybrid infrastructure required for large‑scale quantum applications.