Beaglebone AI¶
The Beaglebone AI is used as secondary processing unit in our RaceCar platform. The board is built around a Texas Instruments (TI) AM5729 system-on-chip (SoC). Its computing capabilities enable the user to develop machine learning applications with ease. It supports all standard communication interfaces over an 46-pin header on either side of the board. Additionally, it also supports a 16-bit LCD interface. The board was developed specifically for AI applications and has an integrated neural engine that processes complex algorithms on the hardware level.
The Beaglebone AI comes with an ARM-A15 based dual-core processor. It supports the standard ARM instruction set with hardware virtualization support. Two dual-core Programmable Real-Tume Units (PRU) subsystems are present to provide ultra low latency. Access to these PRUs are enabled via expansion headers. These dual-core PRUs are based on ARMv7E-M architecture for general purpose usage, particularly real-time control.
Technical Specifications¶
Features |
Specifications |
|---|---|
Maximum Processor Speed |
1.5GHz |
Co-processors |
2xC66x, 4xPRU, 4xARM Cortex-M4, 4xEVE, 2xPowerVR SGX544, IVA-HD |
Analog Pins |
7 (4-wire touchscreen capable) (3.3V) |
System on Chip Main Processor |
Texas Instruments Sitara AM5729, 2x ARM Cortex-A15 |
Digital Pins |
72 (including 7 analog pins which are digital capable) |
Memory |
1GB DDR3, 16GB eMMC flash on-board microSD card slot |
USB |
USB 3.0 Type-C 5Gbps Host/Client Port, USB 2.0 Type-A Host Port |
Network |
1 Gigabit Ethernet, 2.4/5GHz WiFi, Bluetooth 4.2/BLE |
Supported Expansion Interfaces |
4x UART, 16-bit LCD interface, 2x I2C, 2x SPI, PWM, Timer, A/D, CAN, Pulse Capture, Quadrature Encoder, PRU GPIO, etc. |
Known Issues¶
The Beaglebone AI gets heated up with the default OS booted from eMMC storage data. In order to reduce the heat, flash the latest image of the debian OS from here on the SD card. In case the heating issue persists, procure a Beaglebone AI fan cape.