Skip to main content
Breaking News
Tech 6 min read

IBM, Fermilab, and Nvidia Unveil Co-Design Platform for Modular Quantum Computers

Accelerating next-gen quantum-classical supercomputing announced November 16, 2026

IBM, Fermilab, and Nvidia Unveil Co-Design Platform for Modular Quantum Computers
Share Article: WhatsApp Telegram X Facebook LinkedIn
Sponsored Advertisement Google AdSense

On November 16, 2026, IBM, Fermilab, and NVIDIA announced plans to build a modular quantum computing co-design platform that physically co-locates IBM Quantum Nighthawk processors with NVIDIA DGX B200 systems. Hosted with Fermilab's Superconducting Quantum Materials and Systems (SQMS) Center, the initiative targets fault-tolerant scientific computing for high-energy and nuclear physics.

What the IBM, Fermilab, and NVIDIA Modular Quantum Computing Platform Delivers

Unveiled during the Supercomputing 2026 (SC26) news cycle, the collaboration brings together IBM, Fermi National Accelerator Laboratory (Fermilab), and NVIDIA to solve one of quantum engineering's hardest scaling bottlenecks: connecting multiple quantum processors so they operate as a unified, fault-tolerant system alongside GPU supercomputers. Rather than designing quantum hardware and classical high-performance computing (HPC) infrastructure in isolation, the partners are adopting a full-stack co-design approach covering physical interconnects, error-correction decoders, software orchestration, and physics applications.

According to the joint announcement, the initial modular quantum computing platform will physically co-locate IBM quantum computers—including systems powered by IBM Quantum Nighthawk—next to NVIDIA DGX B200 accelerated computing nodes. Nighthawk is IBM's most capable quantum processor to date, engineered to execute circuits of up to 7,500 two-qubit operations across 120 qubits at sampling rates reaching 100,000 samples per second. Placing DGX B200 GPU clusters directly alongside these quantum processing units (QPUs) minimizes communication latency for real-time quantum-classical control loops and error correction.

"Scaling quantum and classical together will open new frontiers in high-energy and nuclear physics and beyond. With Fermilab and NVIDIA, we plan to explore how these real-world problems can help us shape the modular, interconnected architectures that will define the next generation of quantum computers and accelerated classical computing." — Jerry Chow, IBM Fellow and CTO of Quantum-Centric Computing

Architecture Breakdown: Hardware, Interconnects, and Roles

Each partner contributes a distinct hardware or software pillar to the modular quantum computing testbed. Fermilab will coordinate scientific application engagement and researcher access through its HEPCloud infrastructure, laying groundwork for a future U.S. Department of Energy (DOE) quantum computing user facility.

Sponsored Advertisement Google AdSense

Partner Core Hardware & Software Contribution Primary Role in Co-Design Platform

IBM IBM Quantum Nighthawk (120 qubits, up to 7,500 two-qubit ops, 100,000 samples/sec) and quantum networking Quantum processing units, modular interconnect research, and hybrid workflow software

NVIDIA NVIDIA DGX B200 accelerated computing systems and AI/GPU acceleration stack Low-latency classical compute, real-time error-correction decoding, and quantum-classical co-design

Fermilab (SQMS Center) Ultracoherent superconducting radio-frequency (SRF) cavities and HEPCloud infrastructure High-energy and nuclear physics applications, researcher access, quantum memories, and bosonic error correction

Why Modularity and Superconducting Cavities Matter

Scaling a single monolithic quantum chip to tens of thousands of physical qubits introduces severe fabrication, wiring, and crosstalk challenges. Modular quantum computing bypasses these limits by linking smaller, high-fidelity quantum modules over dedicated quantum interconnects. In the platform's opening phase, teams will model and emulate inter-module communication to measure how link errors, latency, and classical feedback loops affect overall circuit fidelity.

  • Physical Interconnect Design: Emulation results will directly guide the SQMS–IBM engineering effort to build optimal physical quantum interconnects between processors.
  • Ultracoherent SRF Cavities: In a subsequent phase, Fermilab's SQMS Center plans to integrate its ultracoherent superconducting radio-frequency cavities into the platform as quantum memories, entanglement buffers, and resources for bosonic error correction.
  • Target Physics Workloads: Researchers will focus specifically on high-energy and nuclear physics (HENP) problems beyond classical supercomputers alone, including lattice gauge theories, neutrino interactions, and complex nuclear simulations.

Native Slurm Integration via the Open-Source QRMI Layer

Alongside the Fermilab co-design announcement, IBM and NVIDIA—working with Pasqal and the broader open-source HPC community—introduced native supercomputer scheduling support for quantum processors through the Quantum Resource Management Interface (QRMI). Historically, Slurm, the dominant workload manager across the world's largest supercomputers, had no native resource abstraction for QPUs, forcing data centers to rely on custom plugins outside the core scheduler.

QRMI bridges this gap as an open-source, vendor-agnostic interface that integrates directly with Slurm Hierarchical Resources (HRES). Because QPUs are shared across multiple classical compute nodes rather than seated inside a single server like GPUs, HRES allows HPC administrators to allocate, reserve, and account for QPU capacity using the same policies already applied to CPUs and GPUs.

  1. Unified Job Submission: Scientists submit hybrid quantum-classical jobs as standard Slurm workloads without separate queues, logins, or custom orchestration layers.
  2. Multi-Vendor Portability: Because QRMI is vendor-agnostic, supercomputing centers can plug in quantum hardware from multiple providers under one consistent scheduling model.
  3. Coordinated Resource Allocation: Administrators enforce fair-share limits across coupled CPU, GPU, and QPU partitions during large-scale sampling runs.

Benchmark Proof: IBM, RIKEN, and NVIDIA Achieve 79× Simulation Accuracy Gain

To illustrate what tightly coupled quantum-classical workflows can already accomplish, IBM, Japan's RIKEN, and NVIDIA published new results using sample-based quantum diagonalization (SQD) to simulate Fe4S4, a strongly correlated iron-sulfur molecular cluster found in biological proteins. Using an intentionally imperfect initial state estimate to steer the quantum processor toward dominant electron configurations, the team executed circuits of up to 4,000 two-qubit operations on 72 qubits and collected 5 million samples.

Sponsored Advertisement Google AdSense

By pairing IBM quantum hardware with NVIDIA GPUs inside RIKEN's ROQUO hybrid supercomputer, the researchers achieved a 79-fold accuracy improvement over prior SQD experiments—landing within 4.3 millihartree of density-matrix renormalization group (DMRG), a gold-standard classical method. Supported by Japan's NEDO, the result underscores why modular quantum computing paired with GPU supercomputers is rapidly transitioning from theoretical research into a practical scientific instrument.

Frequently Asked Questions

When did IBM, Fermilab, and NVIDIA announce the modular quantum computing platform?

The three organizations formally announced plans for the modular quantum computing co-design platform on November 16, 2026, coinciding with SC26 supercomputing announcements.

What hardware powers the initial co-design platform?

The initial deployment co-locates IBM quantum computers—including systems powered by the 120-qubit IBM Quantum Nighthawk processor—with NVIDIA DGX B200 accelerated computing systems, with Fermilab's SQMS superconducting cavities planned for a later phase.

Which scientific problems will the Fermilab co-design platform target?

Fermilab will lead high-energy and nuclear physics (HENP) research on the platform, focusing on computationally demanding simulations such as lattice gauge theories, neutrino interactions, and nuclear structure calculations.

How does Slurm QRMI help supercomputing centers adopt quantum computing?

QRMI integrates natively with Slurm Hierarchical Resources (HRES), enabling HPC centers to schedule, reserve, and manage shared quantum processors alongside CPUs and GPUs within standard Slurm jobs regardless of the QPU vendor.

Sponsored Advertisement Google AdSense
Abdelrahman Shawqi

Abdelrahman Shawqi

Author at Egy Monitor

The manger of EgyMonitor

What do you think of this article? Sign in to react

Comments (0)

Leave a Comment

Be the first to comment on this article!

Related Articles

More from Tech