Several quantum companies, it seems, were busy on the first day of this year’s NVIDIA GTC (GPU Technology Conference) taking place from 16-19 March: an MOU between IonQ and South Korea’s KISTI was signed, as well as a partnership between IQM and Zurich Instruments. Companies PsiQuantum, Pasqal, Atom Computing and Quantum Machines launched new NVIDIA integrations with their hardware or software stacks. Here’s a summary of each announcement:
IonQ and KISTI sign MOU to advance quantum-HPC hybrid technologies in South Korea with NVIDIA NVQLink integration
IonQ has signed a Memorandum of Understanding (MOU) with the Korea Institute of Science and Technology Information (KISTI). The MOU outlines a shared vision for the organizations to explore the advancement of quantum-high performance computing (HPC) hybrid technologies to support the development of a robust ecosystem within South Korea. Specifically, the MOU focuses on seeking to establish a phased approach to integrate IonQ’s quantum hardware with KISTI’s HPC infrastructure using NVIDIA NVQLink.
Other key initiatives include:
Key initiatives anticipated under the MOU include:
- Quantum for HPC and AI: Explore the use of quantum computers to boost the impact of current HPC and ML applications in areas such as logistics, chemistry, material science, and LLM fine tuning.
- Hybrid Infrastructure Integration: Joint research on quantum-HPC hybrid technologies focusing on the integration of IonQ hardware and KISTI HPC infrastructure.
- Advanced Quantum Simulation: Joint research on the simulation of quantum algorithms, applications, and technologies for next-generation hardware.
- AI for Quantum: Development of AI models specifically designed to accelerate quantum technologies.
- Talent and Ecosystem Development: Mutual research visits, workshops, and training to develop local quantum talent and expand the computing ecosystem in Korea.
IQM and Zurich Instruments announce partnership to build real-time quantum error correction demonstrator with NVIDIA NVQLink
IQM Quantum Computers and Zurich Instruments announced a joint project to build and operate a real-time quantum error correction (QEC) demonstrator, enabled by the NVIDIA NVQLink platform. The first demonstrator will combine IQM’s 20-qubit superconducting quantum computer, Zurich Instruments’ new ZQCS Quantum Control System, and GPU-accelerated classical computing enabled by NVQLink. This integrated system supports closed-loop, low-latency decoding, and feedback, to enable capabilities required for operating logical qubits at scale and achieving fault-tolerant quantum computing.
PsiQuantum integrates CUDA-Q with Construct suite
NVIDIA CUDA-Q platform is now available through PsiQuantum’s Construct, its software suite for fault-tolerant quantum application development. The integration enables GPU-accelerated state-vector simulation of large-scale quantum algorithms – Construct users can request early access to GPU-accelerated simulation capabilities within their existing workflows via CUDA-Q. This new integration allows developers to validate and benchmark complex quantum algorithms at scales that were previously impractical when using CPU-based simulators. Integrating CUDA-Q into Construct will provide users with tunable acceleration based on their specific needs, ranging from an 8x speedup to up to 450x.
Pasqal integrates CUDA-Q with its Quantum Resource Management Interface (QRMI) runtime
The integration enables CUDA-Q workloads to be scheduled and orchestrated on Pasqal quantum systems through standard Slurm-based high-performance computing (“HPC”) workflows via QRMI, making quantum processors native accelerators in heterogeneous HPC environments.
Atom Computing integrates NVIDIA NVQLink into proprietary control-systems stack
Atom Computing today announced the successful integration of NVIDIA NVQLink into its proprietary control-systems stack. Through this, Atom Computing hopes to enhance its architecture via:
- Large-scale routing and control of qubits
- Increased logical cycle speeds through accelerated syndrome extraction for quantum error correction
- Deeper integration with the broader NVIDIA CUDA-Q ecosystem for hybrid quantum-classical supercomputing
Quantum Machines launches Open Acceleration Stack, integrating NVIDIA, AMD and Riverlane
Quantum Machines has launched The Open Acceleration Stack, a framework allowing users to integrate any classical processor (XPU) into their quantum control stack. This is essentially an expansion of its Orchestration Platform, the company’s framework for the control and operation of quantum processors. Using its OPNIC (OPX Network Interface Card) and NVIDIA NVQLink, the Open Acceleration Stack links its proprietary Pulse Processing Unit (PPU) and high-performance accelerators. Designed to be open and modular, it facilitates integration with a range of classical processors, from CPUs by companies like AMD, as well as FPGAs, high-performance NVIDIA GPUs, and real-time quantum error correction systems from Quantum Machines’ partners like Riverlane.
