Revolutionizing Quantum Computing through NVIDIA DGX Quantum Program Launch by Quantum Machines

Quantum Machines Unveils the NVIDIA DGX Quantum Early Access Program

In a groundbreaking announcement, Quantum Machines (QM), a leader in quantum control solutions, has introduced the NVIDIA DGX Quantum Early Customer Program. This initiative is part of a collaborative effort with six premier research groups and quantum computer developers. The NVIDIA DGX Quantum represents a milestone in hybrid quantum-classical computing, aiming to enhance quantum research and development through innovative technology.

The scaling of quantum computers significantly increases their dependence on classical resources for vital processes like quantum error correction (QEC) and parameter drift compensation. The NVIDIA DGX Quantum aims to meet this demand by providing the classical processing power needed to drive forward practical quantum supercomputing.

At the heart of NVIDIA DGX Quantum lies the OPX1000, an advanced, modular hybrid control platform. This platform connects seamlessly with the NVIDIA GH200 Grace Hopper Superchips, marking the first instance where accelerated computing is integrated into the quantum computing stack. This integration achieves an impressive low round-trip latency of less than 4 microseconds between quantum control systems and AI supercomputers, significantly outpacing existing solutions.

The launch of the Early Customer Program is currently in motion, engaging key players in the field, including notable academic institutions and national laboratories. Participants include MIT's Engineering Quantum Systems group led by Professor William D. Oliver, the Israeli Quantum Computing Center (IQCC), and quantum hardware developer Diraq, among others.

With its hardware-level integration via PCIe, NVIDIA DGX Quantum effectively alleviates bottlenecks that traditionally hinder quantum computing. It facilitates real-time QEC decoding, AI-driven calibration of quantum processors, and supports high-speed hybrid quantum-classical applications. Additionally, the system is designed to be highly scalable, accommodating more OPX1000 controllers and Grace Hopper servers as needed, thereby ensuring a future-proof infrastructure for accelerated quantum supercomputing.

Dr. Itamar Sivan, CEO and co-founder of Quantum Machines, remarked, "NVIDIA DGX Quantum, equipped with OPX1000, signifies a major transformation in the realm of quantum-classical computing. By merging the leading quantum control technologies with top-tier accelerated computing, we empower quantum computer builders to handle expansive quantum processing units (QPUs). This opens new avenues for quantum researchers, and we eagerly anticipate the advancements and applications that will arise."

Professor Benjamin Huard from École Normale Supérieure de Lyon expressed enthusiasm about the program's potential, stating, "I am thrilled to utilize complex codes for the real-time processing of quantum measurement data in superconducting circuits. The NVIDIA DGX Quantum will empower us to ascertain quantum states and parameter estimations in intricate quantum systems, facilitating groundbreaking physics experiments and improving the fidelity of quantum operations."

Diraq’s Head of Solid-State Theory, Professor Andre Saraiva, highlighted the collaborative effort, saying, "DGX Quantum brings together two of our trusted partners, Quantum Machines and NVIDIA, presenting a product that prepares us for the future of hybrid quantum computing and high-performance computing (HPC). With rapid feedforward capabilities and plug-and-play integration into our existing HPC systems, we are primed to increase our qubit counts and enhance error handling capacities quickly."

Dr. Walter Hahn, a project manager at Fraunhofer IAF, noted the potential of DGX Quantum to expedite hybrid quantum algorithms, thus opening new avenues for further research and innovation.

Furthermore, Dr. Nir Alfasi, general manager at IQCC, emphasized the impressive calibration speeds achieved by the NVIDIA DGX Quantum system, stating, "Installed in the Israeli Quantum Computing Center, this system exhibits record calibration speeds for single and two-qubit gates. Utilizing reinforcement learning agents based on Grace Hopper Superchips, we continuously optimize qubit noise environments, enhancing drive and readout fidelities as a critical step towards effective QEC."

For those interested in further insights, Quantum Machines will showcase the capabilities of NVIDIA DGX Quantum at the APS March Meeting 2025 from March 16-21, at Booth #213 zone A, demonstrating how this innovative platform is fostering the next era of quantum computing. Concurrently, Dr. Ramon Szmuk from Quantum Machines will present a talk titled "Integrated QPU/GPU Calibration of Multiqubit Circuits" at the NVIDIA GTC 2025 on March 19, exploring the advanced integration features of the NVIDIA DGX Quantum system.

About Quantum Machines

Quantum Machines (QM) is a forerunner in providing quantum control solutions, driving the evolution of quantum computing through its Hybrid Control methodology. By uniting quantum and classical operations, Hybrid Control minimizes friction and enhances performance across hardware and software, enabling researchers and builders to iterate rapidly, solve challenges, and actualize visionary concepts. QM’s platform is adaptable to any quantum processor, empowering the industry to scale systems, accelerate breakthroughs, and push the boundaries of what is possible in quantum technology. To learn more about Quantum Machines, visit quantum-machines.co.