http://www.ti.com/general/docs/suppproductinfo.tsp?distId=26&gotoUrl=https%3A%2F%2Fwww.ti.com%2Flit%2Fgpn%2Fam62a3-q1″>View Datasheet

Texas Instruments AM62A/AM62A-Q1 Arm®-Based Processors

Texas Instruments AM62A/AM62A-Q1 Arm®-Based Processors are part of the automotive-grade family of heterogeneous Arm processors. These processors include embedded deep learning (DL), vision processing acceleration and video, display interface, and extensive automotive peripheral and networking options. AM62A/AM62A-Q1 is built for a set of cost-sensitive automotive applications, including in-cabin monitoring systems and drivers, the next generation of eMirror systems, as well as a broad set of industrial applications in building automation, factory automation, robotics, and other markets. The cost-optimized AM62A/AM62A-Q1 provides high-performance computing for both traditional and deep learning algorithms at industry-leading power/performance ratios with a high level of system integration to lower costs and enable scalability for advanced automotive platforms supporting multiple sensor modalities in stand-alone electronic control units (ECUs).

AM62A/AM62A-Q1 contains up to four Arm Cortex®-A53 cores with 64-bit architecture, a vision processing accelerator (VPAC) with image signal processor (ISP) and multiple vision assist accelerators, DL and video accelerators, a Cortex-R5F device management core and a Cortex-R5F MCU channel core. The Cortex-A53s core provides the powerful computing elements necessary for Linux applications and the implementation of traditional vision computing-based algorithms such as driver monitoring. This 7th generation ISP includes flexibility to process a broader sensor suite. It includes RGB-InfraRed (RGB-IR), support for higher bit depth, and features targeting analytics applications. Key cores have the next generation C7000™ DSP from Texas Instruments (“C7x”) with scalar and vector cores. These devices have a dedicated “MMA” deep learning accelerator enabling performance up to 2 TOPS within the lowest power envelope in the industry when operating at the typical automotive worst-case junction temperature of 125°C.

The 3-port Gigabit Ethernet switch has two external ports and one internal port with TSN support and can be used to enable industrial networking options. An extensive peripherals set is included in AM62A/AM62A-Q1 to enable system-level connectivity such as USB, camera interface, MMC/SD, OSPI, CAN-FD, and GPMC for parallel host interface to an external ASIC/FPGA. The AM62A/AM62A-Q1 supports a secure boot for IP protection with the built-in hardware security module (HSM). These devices employ advanced power management support for portable and power-sensitive applications. The Texas Instruments AM62A-Q1 devices are AEC-Q100 qualified for automotive applications.

Features

  • Processor Cores
    • Up to Quad 64-bit Arm Cortex-A53 microprocessor subsystem at up to 1.4GHz
      • Quad-core Cortex-A53 cluster with 512KB L2 shared cache with SECDED ECC
      • Each A53 core has 32KB L1 DCache with SECDED ECC and 32KB L1 ICache with Parity protection
    • Single-core Arm Cortex-R5F at up to 800 MHz, integrated as part of MCU Channel with FFI
      • 32KB ICache, 32KB L1 DCache, and 64KB TCM with SECDED ECC on all memories
      • 512KB SRAM with SECDED ECC
    • Single-core Arm Cortex-R5F at up to 800MHz, integrated to support Device Management
      • 32KB ICache, 32KB L1 DCache, and 64KB TCM with SECDED ECC on all memories
    • Deep Learning Accelerator based on Single-core C7x
      • C7x floating point, up to 40 GFLOPS, 256-bit Vector DSP at 1.0GHz
      • Matrix Multiply Accelerator (MMA), up to 2 TOPS (8b) at 1.0GHz
      • 32KB L1 DCache with SECDED ECC and 64KB L1 ICache with Parity protection
      • 1.25MB of L2 SRAM with SECDED ECC
    • Vision Processing Accelerators (VPAC) with Image Signal Processor (ISP) and multiple vision assist accelerators
      • 315 MPixel/s ISP; up to 5MP @ 60fps
      • Support for 12-bit RGB-IR
      • Support for up to 16-bit input RAW format
      • Line support up to 4096
      • Wide Dynamic Range (WDR), Lens Distortion Correction (LDC), Vision Imaging Subsystem (VISS), and Multi-Scalar (MSC) support
        • Output color format : 8-bits, 12-bits, and YUV 4:2:2, YUV 4:2:0, RGB, HSV/HSL
  •  
  • AEC-Q100 qualified (automotive)
  • Security
    • Secure boot supported
      • Hardware-enforced Root-of-Trust (RoT)
      • Support to switch RoT via backup key
      • Support for takeover protection, IP protection, and anti-rollback protection
    • Trusted Execution Environment (TEE) supported
      • Arm TrustZone-based TEE
      • Extensive firewall support for isolation
      • Secure watchdog/timer/IPC
      • Secure storage support
      • Replay Protected Memory Block (RPMB) support
    • Dedicated Security Controller with user-programmable HSM core and dedicated security DMA and IPC subsystem for isolated processing
    • Cryptographic acceleration supported
      • Session-aware cryptographic engine with the ability to auto-switch key material based on the incoming data stream
        • Supports cryptographic cores
      • AES – 128-/192-/256-Bit key sizes
      • SHA2 – 224-/256-/384-/512-Bit key sizes
      • DRBG with a true random number generator
      • PKA (Public Key Accelerator) to Assist in RSA/ECC processing for secure boot
    • Debugging security
      • Secure software-controlled debug access
      • Security aware debugging
  • Technology / Package
    • 16nm FinFET technology
    • 18mm x 18mm, 0.8mm pitch full-array, 484-pin FCBGA (AMB)

Applications

  • Driver monitoring systems (DMS)/occupancy monitoring systems (OMS)
  • eMirror/camera mirror systems (CMS)
  • Machine vision cameras
  • Barcode scanners
  • Front camera systems
  • Stick-up cameras/video doorbells
  • Autonomous mobile robots (AMR)

Videos

Functional Block Diagram

Block Diagram - Texas Instruments AM62A/AM62A-Q1 Arm®-Based Processors

Texas Instruments AM62A/AM62A-Q1 Arm®-Based Processors