Proposal

Narrative

Scaling a quantum system monolithically is increasingly impractical due to fabrication limits. Distributed Quantum Computing (DQC) has emerged as a scalable alternative by connecting quantum nodes, enabling architectural scalability. For example, companies including IBM, Google, and Cisco are pursuing modular architectures with quantum-classical orchestration. However, there is a critical shortage of quantum educational platforms that allow students and researchers to gain hands-on experience with modular quantum systems. At the same time, the DQC hardware-software co-design remains untested due to the lack of multi-node quantum testbeds. Here, we propose building an HPC-based digital twin for DQC education and research, powered by a multi-GPU (e.g., four H200 NVL GPUs) emulation platform that runs NVIDIA’s quantum frameworks (e.g., CUDA-QX and cuQuantum). On one hand, the infrastructure will serve as an educational resource for quantum courses at FAU, as well as for quantum certificates and quantum events, such as quantum hackathons. On the other hand, it will enable the development of distributed quantum algorithms and techniques. Specifically, we will develop a digital twin infrastructure to emulate quantum circuits across networked quantum modules using a high-performance multi-GPU system (Aim 1). We will then create an educational system on this infrastructure to support both curriculum-based instruction and quantum-related events at FAU. The platform will be open to all students and integrated into coursework (Aim 2). We will advance DQC research by simulating realistic distributed execution of quantum workloads with quantum network constraints using GPU-accelerated quantum simulation backends (Aim 3). Overall, the digital twin will be available to all students for hands-on learning in modular quantum systems, in-class programming for courses leading to quantum certificates and minor programs, as well as experiences through quantum hackathons and laboratories.

Facilities

No physical renovations or upgrades are required. The proposed HPC-based quantum digital twin will be deployed using a multi-GPU server (SuperMicro 5U with 4× NVIDIA H200 NVL GPUs). The server is expected to be installed in FAU’s existing Research Computing/HPC.

Hardware Requirements

The hardware specifications below are provided by Mark III Systems and reviewed by Rhian Resnick (Director, Research Computing, Office of Information Technology, FAU): The configuration of the SuperMicro 4x H200 NVL GPU Server for our quantum computing digital twin: 1. Server chassis and motherboard: OPTIMIZED X14 AP 5U 8GPU GRANITE RAPIDS GEN5 PCIE DUAL ROOT and X14 AP 5U 8GPU GRANITE RAPIDS GEN5 PCIE DUAL ROOT SYSTEM, 2. Two Granite Rapids CPUs: 6960P 2P 72C(2.7G/500W,2.3/400)(66/2.6,48C/3.1/450)432M4DIQD, 3. A total of 768?GB DDR5 ECC memory (32?GB × 24), operating at 6400 MT/s: 32GB DDR5-6400 2Rx8 (16Gb) ECC RDIMM, 4. Two M.2 PCIe Gen4 SSDs: Micron 7450 PRO 960GB NVMe PCIe 4.0 M.2 22x80mm 3D TLC, 5. Two 3.8?TB NVMe SSDs: Micron 7450 PRO 3.8TB NVMe PCIe 4.0 3D TLC U.3 7mm,1DWPD, 6. Mellanox ConnectX-7 2×200?Gb/s dual-port NIC: 900-9X7AH-0078-DTZ std 2x200G EN/IB QSPF112Gen5x16NoCryptCX7, Four NVIDIA H200NVL GPUs: NVIDIA H200NVL, 141GB HBM3e, PCIe 5, 600W, w/ 5yr NVAIE. Please refer to the attached Excel file from Mark III Systems for additional details.

Software Requirements

1. Ubuntu 2. NVIDIA CUDA Toolkit 3. cuQuantum SDK 4. CUDA-Q 5. Qiskit

Personnel Costs

No personnel costs. The hardware installation and software environment setup will be jointly handled by the FAU IT department and EECS faculty or trained student assistants.

Other Costs

No other costs.

Timeline

The project will begin in early 2026 and continue through the summer with the acquisition, installation, and configuration of the hardware, as well as the development of the quantum educational platform. The platform will be piloted in the Fall 2026 semester through integration with quantum computing courses at FAU. Student feedback will be collected and used to iteratively refine the platform. After the pilot phase, the platform will be made accessible to all students engaged in quantum-related coursework or research projects. The platform will be continuously maintained and enhanced throughout the operational lifetime of the hardware.

Sustainability

The proposed platform is designed to have a long-term educational and research impact that extends beyond the initial project period. The installation and system setup will be conducted in collaboration with FAU’s OIT, and Rhian Resnick has reviewed the hardware specification to ensure alignment with institutional infrastructure standards. Following installation, the platform will be maintained and developed jointly by the Department of EECS faculty and trained student assistants. The Department of EECS will take the lead in further developing the platform and actively pitching its capabilities in both internal and external venues to expand its use. The infrastructure will be integrated into the quantum computing curriculum at FAU, supporting quantum-related courses, hackathons, and independent studies. This will ensure recurring use by students and faculty and will grow FAU’s capacity in quantum education and workforce development. If maintenance or upgrades are needed beyond the initial period, the project will seek support from the college and OIT to ensure uninterrupted operation.

Resource Matching

The EECS department will provide faculty oversight and student assistance for installation, system setup, and ongoing platform development. Instructional support will be incorporated into existing teaching assignments for quantum-related courses. We will also seek external resource matching and support letters from industry collaborators (e.g., Cisco, NVIDIA, IBM), who have already expressed interest in distributed quantum computing research and workforce development. These partnerships will lead to external funding proposals.

Implementing Organization

The Department of EECS will lead the implementation of the project, with technical support from OIT.