Altera introduces next-gen FPGAs with integrated ML processing for robotics and factory automation

Altera Corporation introduced new programmable hardware and software for embedded developers working on edge applications. The new offerings include Agilex FPGAs, Quartus Prime Pro software, and FPGA AI Suite for embedded systems development in robotics, factory automation, and medical equipment.

Altera’s programmable hardware addresses technical requirements for embedded and edge applications, focusing on power consumption optimization, performance metrics for edge computing, and form factor considerations. These products integrate with industry-standard frameworks such as TensorFlow and PyTorch, while maintaining compatibility with development tools including OpenVINO and Quartus Prime.

The newly available Agilex 3 FPGAs deliver 1.9x higher fabric performance compared to the previous generation while consuming 38% less power. The FPGAs feature a programmable architecture with built-in AI Tensor blocks and embedded processors, supporting edge and embedded infrastructure with AI capabilities.

For robot control systems, Agilex 3 FPGAs enable real-time control for multi-axis robot arms through sensor fusion with integrated machine learning capabilities. The architecture implements multi-sensor data pipelines to enhance precision and responsiveness in robotic applications.

In smart factory settings, the FPGAs improve defect detection systems through parallel processing combined with convolutional neural networks. This technical approach implements object recognition for comprehensive industrial data analysis, enhancing quality control processes.

Production-quality Agilex 3 devices, development kits, partner boards, and system-on-modules are now available for ordering, allowing embedded system designers to begin implementation and testing for their specific applications.