DEV Community

Cover image for Altera OQC and NVIDIA Unveil Quantum Controller Design
Valentin Podkamennyi
Valentin Podkamennyi

Posted on Originally published at vpodk.com

Altera OQC and NVIDIA Unveil Quantum Controller Design

Altera Corporation, Oxford Quantum Circuits, and NVIDIA announced a collaborative effort to develop an end-to-end quantum system controller reference design. This new architecture utilizes field-programmable gate array technology to link quantum processors with graphics processing units. The project addresses the critical need for compact and efficient control hardware in the quantum sector.

Integrated Hardware for Quantum Control

The collaboration focuses on solving the physical limitations of current quantum computing environments. As quantum processing units gain more qubits, the classical hardware required to control them often grows at an unsustainable rate. Traditional setups require massive racks of radio frequency chains and interconnects that consume significant floor space and power. By using the Agilex 9 Direct RF FPGA, Altera integrates these functions into a much smaller footprint. This specific hardware includes built-in analog-to-digital and digital-to-analog converters.

The integration removes the necessity for external radio frequency up-conversion or down-conversion hardware. This shift allows the control and readout processes to run directly on the chip. Frequency multiplexing is another key feature of this design. It permits multiple readout channels to share a single port. This capability is essential for scaling systems because it prevents the hardware from becoming unmanageable as qubit counts increase. The design effectively reduces the total amount of hardware needed to operate a sophisticated quantum processor.

Farhad Shafai, the head of business solutions and marketing at Altera, noted that the project demonstrates a full system design in a practical setting. The architecture spans from the quantum processor through the FPGA and into the GPU. This setup provides a clear path for developers who want to build their own systems using Direct RF technology. The goal is to provide a stable anchor for the control layer of modern quantum computers. By reducing the physical complexity of the control hardware, the partners are making it easier for research facilities to expand their quantum capabilities without requiring massive infrastructure overhauls.

Synergies Between FPGAs and GPUs

The architecture leverages the unique strengths of both FPGAs and GPUs to manage quantum operations. The FPGA handles real-time qubit control, resonator readout, and spectroscopy tasks. These functions require synchronized scheduling and extremely low latency to maintain the delicate state of the qubits. While the FPGA manages these immediate hardware interactions, NVIDIA GPUs take over the computationally heavy tasks. These tasks include system calibration and machine learning-based classification for readout data.

Connection between these two components happens through NVIDIA NVQLink. This high-speed interface allows for rapid data exchange, ensuring that the heavy processing done by the GPU can influence the real-time operations of the FPGA. CUDA functions drive the logic within the FPGA for various system functions. This hybrid approach ensures that the system is not only fast but also intelligent. It allows for advanced AI-driven calibration techniques that were previously difficult to implement in real-time quantum control environments.

Sam Stanwyck, the director of quantum product at NVIDIA, stated that advances in artificial intelligence are creating new methods for operating quantum computers. The NVQLink provides an open architecture that allows builders to apply these AI advances directly to their hardware. The reference design serves as a functional foundation for developers to explore AI-driven readout and calibration. This collaboration provides a working template that can accelerate the development cycle for new quantum systems. By offloading complex calculations to the GPU, the system maintains high performance while staying flexible enough for future software updates.

Development Tools and Future Availability

The reference design includes a comprehensive software stack and a set of validated functions. These building blocks cover the basics of quantum system control, such as qubit initialization and resonator spectroscopy. Developers also gain access to a GPU-based machine learning classifier. One of the most useful features for engineers is the built-in emulation path. This allows teams to test their algorithms and bring up the controller before they even have physical qubit hardware available. This simulation capability helps teams reach their research goals much faster.

Oxford Quantum Circuits contributes the quantum processing unit and the necessary radio frequency conditioning electronics to the partnership. Simon Phillips, the chief product officer at OQC, emphasized that precise control and fast feedback are mandatory for unlocking the potential of their processors. The programmable nature of the Altera FPGA provides the low-latency connection required for these tasks. As quantum systems evolve, the ability to refine calibration and feedback through software is a significant advantage. This integration simplifies the overall development process and supports a transition toward larger qubit arrays.

The industry will get a first look at this system during Supercomputing 2026 in Chicago. Altera plans to host live demonstrations of the full-system reference design at their exhibition booth. This event provides an opportunity for IT managers and developers to see how the hardware handles real-world quantum workloads. Following the public demonstration, the official release of the Agilex 9 Direct RF quantum system controller reference design is set for January 2027. This timeline gives the industry a clear window for when they can begin integrating these advanced control technologies into their own quantum computing roadmaps.

Top comments (0)