Introduction, features, and applications of AM68 dual-core Arm Cortex-A72 MCU

Texas Instruments' AM68 scalable processor is based on the improved Jacinto 7 architecture and is designed for smart vision cameras and general computing applications. The AM68x series is designed for a variety of cost-sensitive high-performance computing applications in factory automation, building automation and other fields.

AM68 provides high-performance computing technology for traditional and deep learning algorithms with an industry-leading power/performance ratio, with a high level of system integration, scalability and low cost for advanced vision camera applications. Key cores include Arm and GPU processors for general-purpose computing, next-generation DSPs with scalar and vector cores, dedicated deep learning and legacy algorithm accelerators, integrated next-generation imaging subsystems (ISPs), video codecs and isolated MCU island. The core is protected by industrial-grade security and secure hardware accelerators.

The independent dual-core cluster configuration of the Arm Cortex-A72 simplifies multi-OS applications with minimal need for software hypervisors. Up to two Arm Cortex-R5F subsystems support low-level, timing-critical processing tasks, making the Arm Cortex-A72 core unencumbered by applications. Building on existing world-class ISPs, TI's 7th generation ISPs include the flexibility to handle a broader suite of sensors, higher bit depth support, and capabilities for analytics applications. Integrated diagnostics and safety functions support operation up to SIL-2 level, and integrated safety functions protect data from attacks. CSI2.0 port supports multi-sensor input.

The C7000 DSP next-generation core ("C7x") combines TI's industry-leading DSP and EVE cores into a high-performance core and adds floating-point vector computing capabilities to simplify software programming while enabling legacy code backwards compatibility. Dedicated vision hardware accelerators provide vision preprocessing without impacting system performance. The C7x/MMA core can be used for deep learning functions on AM68 class processors.