Quantum Error Correction Platform Links IBM and Oak Ridge
The system pairs IBM's 120-qubit Nighthawk quantum computers with ORNL's Umbra supercomputer, a research-stage platform with no published error-correction benchmark yet.
Updated on Nov. 5, 2026 in Quantum Computing

IBM, AMD, and the Quantum Science Center at Oak Ridge National Laboratory launched a quantum error-correction platform that connects IBM's 120-qubit Nighthawk quantum computers with ORNL's Umbra high-performance computing system. The platform is a co-design system for quantum error correction, announced on November 5, 2026, and is described as a platform rather than a commercially shipped product.
Why it matters
Physical quantum qubits make errors roughly once every few hundred operations, so long or complex calculations cannot be trusted without correction. Pairing quantum hardware with classical supercomputing, and letting researchers measure qubits mid-calculation and act on results in real time, is a step toward error-corrected quantum-classical workflows.
The platform pairs IBM's 120-qubit Nighthawk system, which has square-lattice qubit connectivity, with Umbra, a supercomputer built on AMD Instinct MI355X GPUs and 5th Gen EPYC CPUs. The announced figures describe the hardware configuration; no integrated error-correction benchmark was reported.
The players
IBM
IBM builds superconducting quantum computers, including the 120-qubit Nighthawk system used in this platform.
AMD
AMD supplies the Instinct MI355X GPUs and 5th Gen EPYC CPUs that power the Umbra system.
Oak Ridge National Laboratory
Oak Ridge National Laboratory, a U.S. Department of Energy science facility, operates the Umbra high-performance computing system and co-leads the Quantum Science Center partnership.
The details
Quantum bits, or qubits, store information in fragile physical states that drift when disturbed by noise, which is why errors accumulate. Quantum error correction spreads one logical bit of information across many physical qubits and uses precisely timed control signals to detect and fix errors without destroying the stored data. The platform lets researchers set exact timing for those protective signals and read qubit states mid-calculation, then feed the results back to classical processors on Umbra in real time.
Timeline
Earlier in 2026, IBM and ORNL showed that quantum, classical, and Frontier supercomputer workloads could simulate magnetic materials.
Earlier in 2026, IBM and ORNL calculated molecular configurations of a promising fusion fuel material.
On November 5, 2026, IBM, AMD, and ORNL announced the quantum error-correction platform.
The Tech Race
The platform follows IBM and ORNL's earlier 2026 demonstrations of magnetic-material and fusion-fuel molecular simulations on combined quantum and supercomputing resources. It positions error correction as the next bottleneck in the competition to make quantum hardware useful for scientific computing.
The platform is a research infrastructure announcement, so no consumer availability, pricing, or access program was reported. Researchers and institutions working on quantum error correction are the group most likely to engage with it first.
The takeaway
The result shows that scientific quantum computing now depends as much on classical supercomputing and real-time control as on qubit count. Watch for published error-correction benchmarks from the IBM, AMD, and ORNL platform, which would show whether it reduces the errors that currently limit long calculations.
Further reading
For broader context on how these machines are being built and tested, explore our Quantum Computing coverage.





