Altera today unveiled an FPGA-based quantum system controller reference design co-developed with Oxford Quantum Circuits (OQC) and NVIDIA. An Altera Agilex 9 Direct RF FPGA carries out real-time control for a next-generation OQC quantum processing unit (QPU) via RF conditioning electronics engineered by OQC, while NVIDIA NVQLink links the FPGA to NVIDIA accelerated compute hardware. Altera plans to release the design in January 2027 and is showcasing the complete system at SC26 in Chicago.
![]()
FPGA Control and GPU Processing
Altera’s design is built around higher hardware density, as qubit counts are expanding faster than the physical space allocated for classical control hardware surrounding quantum processors. FPGAs consolidate functions that would otherwise require racks of separate RF chains, converters, and interconnects. The Agilex 9 Direct RF FPGA embeds analog-to-digital and digital-to-analog converters, enabling control and readout operations without external RF up-conversion or down-conversion circuitry.
Altera cites direct RF sampling speeds up to 64 GSPS for the Agilex 9 Direct RF-Series, though it has not specified which chip variant the reference design employs. Frequency multiplexing allows multiple readout channels to share a single RF port, which Altera states cuts down on extra RF hardware required as qubit scales grow. Within this reference design, the FPGA manages real-time qubit control, resonator readout, and spectroscopy, with synchronized scheduling across these tasks.
NVQLink, NVIDIA’s open framework for connecting QPUs with GPU systems, bridges the FPGA to GPUs that handle compute-intensive workloads such as calibration and machine learning-powered readout classification. CUDA routines in turn steer FPGA logic to execute QPU system operations. Other quantum hardware developers, including Quantum Machines, are also advancing related quantum-classical integration work.
Farhad Shafai, head of business solutions and marketing at Altera, noted the demonstration “gives developers a clear path to build on.” Simon Phillips, CPO at OQC, commented that the FPGA delivers a programmable, low-latency interface to its QPU, uniting control and readout functions while retaining flexibility to fine-tune calibration and feedback loops.
Software Stack and Emulation
Developers receive a validated suite of controller functions plus a software stack covering qubit and resonator spectroscopy, qubit initialization, and resonator readout and classification. The package also includes a GPU-based machine learning readout classifier and sample GPU initialization routines for the QPU. Sam Stanwyck, director of quantum product at NVIDIA, described it as a functional reference design for investigating AI-driven calibration and readout techniques.
An integrated emulation flow lets teams bring up the controller and validate algorithms before physical qubit hardware becomes available. Altera says this methodology supports developers in building smaller, lower-power, lower-latency controllers, though it has not disclosed latency, power draw, or supported qubit count figures for the design. The live demo is running at Altera booth #2854 during SC26.
Beijing Qianxing Jietong Technology Co., Ltd.
Sandy Yang/Global Strategy Director
WhatsApp / WeChat: +86 13426366826
Email: yangyd@qianxingdata.com
Website: www.qianxingdata.com/www.storagesserver.com
Business Focus:
ICT Product Distribution/System Integration & Services/Infrastructure Solutions
With 20+ years of IT distribution experience, we partner with leading global brands to deliver reliable products and professional services.
“Using Technology to Build an Intelligent World”Your Trusted ICT Product Service Provider!



