AMD Accelerates Hybrid Quantum-Supercomputing Innovation 

AMD is driving the next era of computing by laying the foundation for scalable, hybrid systems that combine classical HPC, AI, and quantum processing. As quantum technology advances from research to early-stage deployment, AMD delivers the high-performance CPUs, GPUs, FPGAs, and networking needed to build practical quantum compute platforms. 

From data centers to research labs, AMD high-performance technologies are helping enable a new class of compute systems—designed to address challenges that exceed the capabilities of classical computing.

AMD Vision for Hybrid Quantum-Supercomputing Architectures 

The future of computing is a hybrid, quantum-accelerate supercomputing architecture. To unlock the full value of quantum systems, AMD is focused on co-developing platforms that tightly integrate high-performance computing and AI components—including AMD EPYC™ processors, AMD Instinct™ accelerators, Versal™ adaptive SoCs, AMD Pensando™ AI NICs and networking solutions—with quantum hardware and open software stacks. 

AMD is working with academic institutions, national labs, and enterprise partners to build out the architectural blueprints, software frameworks, and system-level innovations required for Quantum-HPC-AI convergence. 

Collaboration with IBM: A Key Proof Point 

A recently announced strategic partnership between AMD and IBM demonstrates this commitment. Through the collaboration, the two companies will co-design proof-of-concept systems that link IBM quantum systems with AMD compute engines. 

This collaboration is focused on the joint development of: 

• Hybrid architectures where CPUs, GPUs, FPGAs, and QPUs operate together 

• Workflows that showcase early-stage quantum utility 

• Open-source software to support industry-wide adoption 

• Adaptive compute solutions that manage quantum I/O, control, and error correction  

The IBM collaboration builds on a broader AMD strategy to accelerate innovation across classical and quantum domains, supporting scalable solutions that can evolve as quantum hardware matures. 

AMD Technology Powering Quantum Research 

AMD platforms are powering quantum exploration at top research institutions worldwide. The AMD compute portfolio addresses multiple needs across data processing, analysis, networking, and real-time quantum control.  AMD EPYC™ CPUs orchestrate data preparation while  AMD Instinct™ GPUs accelerate massive data processing and AI-enabled analysis on quantum outputs. AMD FPGAs and adaptive SoCs provide the interface with quantum systems and offer promise for enabling real-time error correction to maintain application integrity. AMD Pensando™ Pollara 400 AI NIC is the world's first Ultra Ethernet Consortium-ready AI NIC that brings programmable congestion control to large- scale hybrid computing systems.

 Notable collaborations include:

• The LUMI supercomputer, one of the most powerful and energy-efficient computing systems in the world, powered by AMD is enabling the exploration of HPC-quantum integration.  

• On the LUMI supercomputer, researchers used AMD GPUs to run quantum chemistry simulations using PennyLane’s VQE (Variational Quantum Eigensolver) workflows.

• Quantum Machines, specializing in quantum control systems, is using AMD FPGAs for exceptional real-time quantum control performance.

• Riverlane leverages AMD adaptive technology using Zynq™ SoCs to innovate operating systems for quantum computers.

• The AMD University Program launched quantum summer school in partnership with the Fermi National Accelerator Laboratory to support the use of AMD RFSoC development systems in the summer school program hosted by the U.S. Quantum Information Science center at Oak Ridge National Laboratory.

These capabilities position AMD at the heart of today’s quantum-compute research and development efforts—from low-latency signal processing to high-throughput classical compute.

Enabling Open Ecosystems for Hybrid Computing 

To support rapid innovation and broad adoption, AMD champions open standards and open-source tools, including ROCm™ software, contributions to PyTorch, vLLM, SGLang, QICK  and other AI, HPC, and quantum computing frameworks. AMD continues to invest in software integration layers that allow developers to efficiently build and scale hybrid quantum-classical workflows. 

 AMD actively contributes to and supports open-source projects that intersect with quantum computing, like Linux® Kernel, PyTorch, TensorFlow, JAX, and Hugging Face. These frameworks are increasingly used in quantum machine learning and hybrid quantum-AI workflows.

AMD and the Open Compute Project (OCP) are collaborating to shape the future of computing infrastructure.  As a founding member of the Ultra Ethernet Consortium (UEC), AMD collaborates with ecosystem partners to advance networking for AI, HPC, and quantum workloads. It aims to redefine Ethernet for AI and HPC workloads, addressing latency, bandwidth, and scalability gaps, to scale AI clusters across racks with supercomputing-grade performance while maintaining backward compatibility.

UALink™ (Ultra Accelerator Link) is an open industry standard for high-performance accelerated compute scale-up communication, enabling unified compute fabrics. The open approach will allow quantum hardware vendors to adopt UALink for modular, scalable quantum systems.

By expanding access to open software tools, open architectures, and open networking solutions, AMD is enabling  a sustainable hybrid computing ecosystem.

Looking Ahead: From Research to Real-World Utility 

AMD work in hybrid quantum-classical computing is designed to drive measurable performance gains across workloads that demand massive parallelism, agentic AI, advanced modeling, and quantum computing. These include drug discovery, cryptography, supply chain optimization, and machine learning.  AMD collaboration with leading labs like Oak Ridge National Laboratory are a catalyst to accelerate discovery and innovation. The Oak Ridge Quantum Computing User Program (QCUP) provides access to state-of-the-art quantum computing resources. Future demonstrations will aim to showcase  the convergence of HPC, AI, and quantum computing to address the world’s most challenging problems, paving the way for new sovereign compute paradigms and broader deployment of quantum-accelerated computing.  Please visit the AMD Quantum Computing page to keep apprised of the latest advancements from AMD.