The AM68Ax features a system-on-a-chip (SoC) architecture with extensive integration, providing high-performance computing technology for both traditional and deep learning vision applications. With industry-leading power and dual 64-bit Arm® Cortex®-A72 microprocessors and GPU processors running at 2 MHz with high-level system integration, the AM68Ax enables scalability for advanced vision applications with next-generation digital signal processing (DSP) featuring multiscale and vector cores, and dedicated deep learning accelerators delivering 1 to 32 teraoperations per second (TOPS).
The AM68Ax features an integrated, next-generation image capture subsystem (ISP) that supports multi-stream video codecs and an isolated microcontroller island powered by a dual-core Arm Cortex-R5F MCU, with secure multi-display outputs protected by industrial-grade hardware safety and security accelerators. The AM68Ax is fully compatible with open-source modeling and evaluation tools to help designers develop and scale edge AI designs across multiple systems.
