Skip to content
Innovation

IBM, AMD and Oak Ridge Launch Quantum Error Correction Platform

IBM, AMD and Oak Ridge National Laboratory launch a platform combining quantum computing and high-performance computing for error-correction research.

IBM, AMD and Oak Ridge National Laboratory have launched a quantum error correction platform that combines quantum processors with high-performance computing. The initiative aims to help researchers develop and test methods for making quantum computers more reliable.

The first deployment brings IBM's 120-qubit Quantum Nighthawk processor together with Umbra, a high-performance computing system operated by Oak Ridge's Quantum Science Center. Umbra uses AMD Instinct MI355X GPUs and fifth-generation AMD EPYC CPUs to provide the classical computing power needed for error-correction research.

The platform is located at IBM's facility in Poughkeepsie, New York. It is designed to give researchers access to closely integrated quantum and classical computing resources without requiring them to build the infrastructure themselves.

How the Quantum Error Correction Platform Works

Quantum computers use qubits to process information. Unlike conventional computer bits, qubits are highly sensitive to environmental disturbances and other sources of noise. These effects can introduce errors during calculations and limit the reliability of quantum systems.

Quantum error correction aims to detect and manage those errors. It uses groups of physical qubits to protect information, while classical computing systems analyse measurement results to identify potential faults. The software responsible for this process is known as a decoder.

The decoder must process information quickly enough to keep pace with the quantum processor. This makes powerful classical computing an important part of developing practical error-correction methods.

By placing IBM's quantum hardware alongside Oak Ridge's Umbra system, the new platform allows researchers to test how GPUs and other computing resources can support decoding, optimisation and software coordination. It also reduces the delays and complexity that can arise when quantum processors and supercomputers operate in separate locations.

IBM, AMD and Oak Ridge Target More Reliable Quantum Systems

The collaboration builds on existing work between the three organisations. IBM and Oak Ridge have previously explored workloads that combine quantum computers with the Frontier supercomputer. IBM and AMD have also worked together on using AMD GPUs for quantum computing tasks, including research into quantum error decoding.

The new platform expands that cooperation by giving researchers an environment to evaluate different error-correction approaches on actual quantum hardware. Teams can compare the computing resources, processing time and number of qubits required by different methods.

The initiative also supports the US Department of Energy's Quantum Genesis Mission. Its Q Competition targets a system capable of operating 100 logical qubits through 100,000 demanding operations by late 2028.

IBM says its current experiments have reached up to 70 logical qubits and 438 such operations. The company estimates that reaching the competition's target will require an approximately 230-fold increase in the number of operations.

A Step Towards Fault-Tolerant Quantum Computing

The platform is intended to help researchers understand what future quantum systems will need to operate reliably at a larger scale. Its work will cover error-correction techniques, GPU-based decoding and the software needed to coordinate quantum processors with classical computing infrastructure.

The research may also inform IBM's development plans for Quantum Starling, its proposed fault-tolerant quantum computer, which the company aims to deliver by 2029.

However, the new platform does not mean that practical fault-tolerant quantum computing has been achieved. It provides a research environment for testing the hardware and software approaches that could help make that goal possible.

By combining IBM's quantum processors with Oak Ridge's supercomputing infrastructure and AMD's computing hardware, the partners are addressing one of the central challenges facing the field: making quantum calculations more dependable as systems become more capable.