Andre Saraiva, Head of Theory at Diraq, highlighted a crucial advancement in quantum computing integration with the implementation of NVIDIA NVQLink. NVIDIA has addressed this need with NVQLink, superseding the earlier DGX Quantum and providing an open platform for orchestrating Diraq’s quantum processing unit (QPU) and its control hardware. Saraiva envisions NVQLink as a holistic platform, unlocking collaboration opportunities and accelerating Diraq’s operations significantly. However, realizing this potential requires a low-latency connection between the control electronics for qubits and the high-performance computers needed for error correction and hybrid applications. NVIDIA’s NVQLink addresses this critical need, providing a solution Diraq’s engineers have been exploring since initial work with NVIDIA DGX Quantum earlier this year, according to the company.
NVQLink supersedes DGX Quantum, offering an open platform for orchestrating multiple aspects of Diraq’s quantum processing unit (QPU) and its associated hardware. Speaking at NVIDIA AI Day in Sydney, Diraq’s Head of Theory, Andre Saraiva, emphasized the importance of this low-latency link for the company’s system. Saraiva also highlighted the collaborative benefits of NVQLink, noting that NVIDIA’s widespread adoption and powerful systems will encourage broader engagement and experimentation within the quantum computing community.
Building on this, Diraq is collaborating with the Quantum Performance Lab at Sandia National Labs to leverage NVQLink for real-time QPU calibration. This work will involve developing techniques ranging from lightweight protocols to robust reinforcement learning policies, utilizing both the computational power of NVQLink and its tight integration with Diraq’s QPU. Furthermore, Diraq plans to use NVQLink to characterize noise within its system and build machine-learning models that will inform future error-correcting codes and useful quantum algorithms, integrating NVIDIA CUDA-Q into its software development process.
Diraq & NVIDIA Forge New Path for Scalable Quantum Systems
Diraq is leveraging the NVIDIA NVQLink platform to enhance real-time calibration of its quantum processing unit (QPU). According to the company, this integration allows for the development and implementation of techniques ranging from lightweight protocols to more robust, reinforcement learning-based policies. The collaboration with Sandia National Labs is central to this effort, combining Diraq’s hardware with Sandia’s expertise in quantum control.
Building on this, Diraq intends to utilize NVQLink for comprehensive system characterization, specifically focusing on noise analysis. The company plans to construct a machine-learning model informed by this data, ultimately refining the error-correcting codes employed in its quantum algorithms. This process will be facilitated by NVIDIA CUDA-Q, NVIDIA’s open-source platform for hybrid quantum-classical application development, providing a cohesive software environment. This approach will enable Diraq to proactively address and mitigate sources of error within its QPU.
Andre Saraiva, Head of Theory at Diraq, highlights the collaborative benefits of the NVQLink platform. Saraiva emphasizes that NVQLink unlocks opportunities to work with other teams, as NVIDIA systems are widespread and familiar to many researchers. This widespread adoption fosters innovation and allows for easier experimentation with new ideas within the quantum computing community. The platform’s accessibility, combined with its powerful computational capabilities, positions NVQLink as a key enabler for future advancements in scalable quantum systems.
“It took us less than a week to find four ways that DGX Quantum would substantively accelerate our operations. In comparison, NVQLink is a far more mature technology, and its potential is unbounded.”
Head of Theory Andre Saraiva, Diraq
“This platform unlocks opportunities to work with other teams, because NVIDIA systems are so powerful and so widespread. People get excited about having an interface that they are familiar with, and that they can use to try out new ideas. As instruments for collaboration in quantum computing, NVIDIA tools are second to none.”
Head of Theory Andre Saraiva, Diraq
The integration of NVIDIA NVQLink with Diraq’s silicon qubits represents a significant step towards scalable quantum computing systems. This low-latency link enables Diraq to fully harness the speed of its qubits, facilitating rapid error correction and complex hybrid computations.
“Integration between quantum processors and accelerated computing is key for running complex control tasks and deploying hybrid applications”
General Manager for Quantum Tim Costa, NVIDIA
For industries reliant on high-performance computing, this development could enable advancements in areas demanding immense processing power. As Andre Saraiva of Diraq highlighted, the widespread adoption of NVIDIA systems fosters collaboration and accelerates innovation. The implications extend beyond quantum computing itself, offering a familiar and powerful platform for researchers and developers alike.
