AMD has unveiled the second-generation Versal adaptive systems on chip, which aims to deliver end-to-end acceleration in embedded systems with a unified approach to computing resources. The new device eliminates the need for multi-chip solutions by combining all computing resources on a single chip, offering an optimal combination of processors for preprocessing, AI inference, and post-processing.
The Versal AI Edge Series Gen 2 features an FPGA programmable logic structure for real-time preprocessing and an array of vector processors for efficient AI inference. Additionally, CPU cores provide the post-processing power needed for complex decision-making and control of safety-critical applications. This design streamlines the acceleration process and improves overall efficiency in embedded systems.
Furthermore, AMD has introduced the Versal Prime Series Gen 2, targeting integrated systems. This series combines programmable logic for sensor processing with high-performance integrated Arm CPUs, offering up to ten times more scalar computing power than the first generation. It also provides more efficient management and supports video signal processing for up to 8K in multi-channel workloads.
In conclusion, AMD’s latest Versal chip series are designed to enhance the performance and efficiency of embedded systems, offering a comprehensive solution for various applications. The unified approach to computing resources and the combination of processors for different phases of acceleration make these chips a promising advancement in the field of artificial intelligence-driven systems.
