Altera announced the release of FPGA AI Suite 26.1.1, a major update to its AI software platform. FPGA AI Suite is designed to simplify and accelerate deployment of trained AI models onto FPGAs for edge AI applications powering physical AI systems including robotics and real-time autonomous machines. The 26.1.1 release introduces a new compiler technology that uses spatial mapping of AI models. This approach delivers ASIC-like performance for optimized AI inference, while maintaining fast time to market and re-programmability for evolving workloads, along with deterministic, low-latency FPGA-based execution.
The spatial compiler architecture in FPGA AI Suite 26.1.1 maps neural networks directly onto FPGA hardware, enabling a streaming dataflow approach instead of traditional sequential processing. By optimizing data movement and parallel execution, it improves efficiency and delivers higher throughput, lower power consumption, and deterministic low latency. This makes it well suited for real-time edge applications such as vision, video analytics, language models, and sensor processing in physical AI systems.
Edge AI is redefining how physical AI systems, such as robotics and autonomous machines, interact with the real world, enabling them to sense, think, and act in real time across dynamic environments. These physical AI systems demand high-performance, low-latency sensor processing, efficient AI inference, and deterministic responses to enable safe and intelligent operation.
Altera supports these applications with its AI-enabled Agilex FPGAs and FPGA AI Suite, delivering deterministic, low-latency performance with the flexibility to adapt to evolving AI workloads. By combining adaptable hardware with optimized AI development tools, Altera enables developers to build and deploy scalable, real-time edge AI solutions across its Agilex portfolio. Agilex FPGAs offer a range of capacities, peripherals, and interfaces to support edge AI applications across different form factors. These solutions enable next-generation physical AI systems with built-in safety, security, and reliability.
About FPGA AI Suite
FPGA AI Suite is a software platform that simplifies AI development and deployment on FPGA-based systems. Developers can use FPGA AI Suite in conjunction with industry-standard frameworks like PyTorch and TensorFlow and the model optimization capabilities from OpenVINO. FPGA AI Suite enables developers to efficiently optimize and run AI models with low latency and deterministic performance. Designed for flexibility and scale, FPGA AI Suite supports applications from edge to data center, helping organizations accelerate time-to-market for real-time, power-efficient AI solutions.
For a complete list of FPGA AI Suite 26.1.1 features, visit what’s new in FPGA AI Suite.
Availability
FPGA AI Suite 26.1.1 is available for download today and supports Quartus Prime Pro Edition 26.1. FPGA AI Suite continues to support early-stage development with license-free operation for up to 100,000 consecutive inferences, lowering barriers to adoption.
For more information, visit altera.com.