Development board speeds designs for AI at the Edge

The BeagleBoard.org Foundation announces general availability of the BeagleBoard.org BeagleBone low-cost development board enabling new AI applications. Built on our proven open source Linux approach, BeagleBone AI fills the gap between small single board computers (SBCs) and more powerful industrial computers. Leveraging the Texas Instruments Sitara AM5729 processor, developers have access to powerful machine learning capabilities with the ease of the BeagleBone Black header and mechanical compatibility.

BeagleBone AI makes it easy to explore how artificial intelligence (AI) and machine learning can be used in everyday life. Through BeagleBone AI, developers can take advantage of the TI C66x digital-signal-processor (DSP) cores and embedded-vision-engine (EVE) cores on the Sitara AM5729 processor. Machine learning is supported by an optimized TI Deep Learning (TIDL) software framework and pre-installed software tools. Downloading and running the latest examples of edge inference algorithms and applications is only a few clicks away.

“This board is the answer to our community’s request to see the next major advancement in the BeagleBone® family,” says Jason Kridner, co-founder of the BeagleBoard.org Foundation. “Its feature set is jam-packed and offers capabilities unparalleled by any other single board computer, open hardware or not.”

BeagleBone AI has a feature set that includes:

BeagleBone Black mechanical and header compatibility
- Works with existing BeagleBoard.org BeagleBone® cape add-on boards and many available third-party cape add-on boards and enclosures
1GB RAM and 16GB on-board eMMC flash with a high-speed interface
USB Type-C for power and superspeed dual-role controller; and USB Type-A host
Gigabit Ethernet, 2.4/5GHz WiFi, and Bluetooth 4.2/BLE
microHDMI video and audio output
Zero-download out-of-box software experience
TI Sitara AM5729 processor with:
- Dual 1.5GHz Arm® Cortex®-A15 microprocessor subsystem
- 2 C66x Floating-Point VLIW DSPs
- 4 Embedded Vision Engines (EVEs) supported by the TIDL software framework library for machine learning
- 2x Dual-Core Programmable Real-Time Unit (PRU) subsystems (4 PRUs total) for ultra-low latency control and software generated peripherals
- 2x Dual Arm® Cortex®-M4 co-processors for real-time control and subsystem management
- IVA-HD subsystem with support for 4K at 15fps H.264 encode/decode and other codecs at 1080p60
- Vivante® GC320 2D graphics accelerator
- Dual-Core PowerVR® SGX544™ 3D GPU