Dev48
Language
  • About
  • Services
  • Industries
  • Technologies
  • Articles
  • Contacts
Book a call
    Home/Articles/Ibm amd advance quantum error correction decoding at scale
Dev48
info@dev48.org+375 29 320-40-96WhatsApp@isky_by

© 2026 · All rights reserved.

IBM, AMD advance quantum error correction decoding at scale

Source: IBM•November 16, 2026 at 04:49 PM

IBM, AMD advance quantum error correction decoding at scale

New open-source tool from IBM and AMD helps researchers evaluate quantum error-correction decoders at scale for future fault-tolerant quantum systems.

Key takeaways:

  • IBM and AMD released Pelagos, a new open-source framework for evaluating quantum error-correction decoders at scale.
  • Pelagos uses GPU acceleration to help researchers run large decoder-evaluation campaigns more efficiently across modern computing infrastructure.
  • New research demonstrates efficient multi-GPU scaling for Pelagos on AMD Instinct™ accelerators and ORNL HPC systems.
  • An accompanying survey paper examines the growing role of decoding systems, classical infrastructure, and hardware-software co-design in fault-tolerant quantum computing.
  • Together, these efforts support the IBM-AMD vision for quantum-centric supercomputing and the recently launched quantum error-correction co-design platform.

Earlier this month, Oak Ridge National Laboratory (ORNL), IBM, and AMD announced a new quantum error-correction co-design platform that brings together IBM quantum processors and AMD-powered HPC infrastructure. At SC26, the collaboration is taking another step toward that vision with the open-source release of Pelagos, a GPU-accelerated framework designed to help researchers evaluate quantum error-correction decoders at scale.

Dealing with quantum computers means dealing with quantum errors. To make quantum computing useful at scale, we must detect and correct those errors continuously—and much of the computational work surrounding that process happens on classical computers. As quantum error correction advances, proving that a decoder works becomes increasingly computationally demanding. Researchers often need to evaluate billions of sampled error patterns to understand how a decoder performs under realistic conditions. Those studies can require large simulation campaigns and substantial classical computing resources. Pelagos was built to make that work faster, more scalable, and more accessible to the broader research community.

Alongside the Pelagos release, IBM and AMD are publishing new research demonstrating how the framework scales efficiently across multiple GPUs, including deployments on AMD-powered HPC systems.

An accompanying survey paper places these developments in a broader context, highlighting the growing importance of classical infrastructure, decoding systems, and hardware-software co-design in the future of fault-tolerant quantum computing. Its central message is that fault-tolerant quantum computing will require advances not just in quantum hardware, but also in the classical systems responsible for decoding, orchestration, and data movement.

The path to fault-tolerant quantum computing is an enormous classical-computing challenge, requiring CPUs, GPUs and sophisticated software to process, evaluate and ultimately help correct errors at tremendous scale. That's where IBM and AMD see an opportunity to bring together leadership quantum technology and high-performance classical computing.

Accelerating decoder research with Pelagos

Quantum computers are inherently susceptible to errors, making quantum error correction a foundational requirement for useful large-scale quantum computing. Error-correcting codes protect quantum information by continuously checking for signs that an error has occurred. Those checks generate data known as syndromes, which identify where an error may have happened without revealing the underlying quantum information itself.

A decoder analyzes those syndromes and determines the most likely correction. As codes grow larger and researchers push toward lower logical error rates, evaluating decoder performance requires increasingly large simulation campaigns and substantial classical computing resources. Pelagos, named after the Greek word for “open sea,” is designed to help researchers navigate the challenge of decoder evaluation. Developing better decoders is essential for reliable quantum computing, but evaluating them can be extraordinarily expensive.

Pelagos accelerates this workflow by combining GPU acceleration with a developer-friendly software architecture. Built using the Triton programming framework, it supports both AMD and NVIDIA GPU environments while providing a common platform for large-scale decoder evaluation. Built as an open-source framework, Pelagos gives researchers an approachable, reusable foundation for developing and evaluating new decoder algorithms.

The result is a tool designed for quantum error-correction researchers, decoder developers, and HPC practitioners alike. Earlier studies showed Pelagos delivering up to 97× higher throughput than a single CPU core and performance comparable to hundreds of CPU cores for decoder-evaluation workloads. By making the framework open source, IBM and AMD aim to lower barriers to experimentation and encourage broader collaboration across the community.

The release is accompanied by new research focused on scaling Pelagos across modern multi-GPU systems. While Pelagos already enabled large distributed studies, the new work improves how individual compute nodes coordinate and utilize multiple GPUs. The approach uses process-level orchestration, parallel host-side operations, and shared-memory techniques to maximize throughput while minimizing coordination overhead.

Tests on AMD Instinct™ MI300X and MI355X accelerators demonstrated strong scaling across eight GPUs, with the GPU decoding stage achieving speedups of up to roughly 7.5× relative to a single-GPU configuration. Researchers also deployed the framework on Lux, an AMD-powered system at ORNL, illustrating how Pelagos can take advantage of production-class HPC infrastructure.

Scaling quantum error correction for the HPC era

For the HPC community, these results show how modern GPU infrastructure can accelerate decoder-evaluation studies as quantum error-correction codes and decoders become more sophisticated. As researchers investigate larger codes, more sophisticated decoders, and lower target error rates, the classical resources required to evaluate those systems continue to grow. Efficient use of GPUs and large-scale compute environments will play an increasingly important role in advancing the field.

That broader trend is explored in the accompanying survey paper on hierarchical decoding for quantum error correction. The paper reviews the rapidly expanding landscape of decoding algorithms, execution strategies, and hardware platforms, from CPUs and GPUs to FPGAs and specialized accelerators.

Viewed in that context, Pelagos becomes more than a software release. It is a practical example of the kind of infrastructure the field increasingly needs. The survey paper provides a framework for understanding the growing demands of decoding systems, while Pelagos gives researchers a concrete tool they can use today to explore solutions.

The project also reflects the broader direction of IBM and AMD's collaboration. The recently launched ORNL-IBM-AMD quantum error-correction co-design platform combines IBM quantum hardware with AMD CPUs and GPUs in a shared environment optimized for low-latency quantum-classical workflows. Pelagos complements that effort by helping researchers develop and validate the decoding techniques that future error-corrected systems will depend upon.

Building the infrastructure for fault-tolerant quantum computing

As the quantum industry moves toward larger logical systems and increasingly sophisticated error-correction strategies, collaboration between quantum and classical computing communities will become even more important. Pelagos is designed to help bridge those worlds by bringing modern HPC techniques to one of the most important challenges in fault-tolerant quantum computing.

Researchers can explore the Pelagos project on , review the new scaling paper, and read the survey paper to learn more. IBM and AMD welcome feedback, contributions, and collaboration from the broader quantum computing and HPC communities as they continue advancing the infrastructure needed for the next generation of quantum computing research.

← All articles

What this article says

More in Software Development

All →
Euler meets Codd
Neo4j

Euler meets Codd

Ansys LS-DYNA Runs Faster on the NVIDIA Vera CPU
Ansys

Ansys LS-DYNA Runs Faster on the NVIDIA Vera CPU

SC26: IBM shows progress in bringing quantum & HPC together
IBM

SC26: IBM shows progress in bringing quantum & HPC together

Translations are now generally available
Medusajs

Translations are now generally available

2027 Ocean Market Outlook
Freightos

2027 Ocean Market Outlook

Turnitin Student Data & AI: Facts About How We Use Your Data
Turnitin

Turnitin Student Data & AI: Facts About How We Use Your Data

More from IBM

SC26: IBM shows progress in bringing quantum & HPC together
IBM

SC26: IBM shows progress in bringing quantum & HPC together

What is a spin qubit?
IBM

What is a spin qubit?

QDC 2026: what quantum developers need to know
IBM

QDC 2026: what quantum developers need to know

Tracking progress on the path to IBM Quantum Starling
IBM

Tracking progress on the path to IBM Quantum Starling

The year of advantage—and what comes next
IBM

The year of advantage—and what comes next

Building policy enforcement into the AI stack with Praxis
IBM

Building policy enforcement into the AI stack with Praxis