Specifications
Product Family
TI SK-AM69, SK-AM68 and SK-AM62A starter kits (evaluation modules)
SK-AM69 Processor
AM69A octa-core Arm Cortex-A72 with deep-learning accelerator, up to 32 TOPS
SK-AM68 Processor
AM68A with two 64-bit Arm Cortex-A72 cores and an 8 TOPS AI accelerator
SK-AM62A Processor
AM62A low-power vision processor, up to 5 MP at 60 fps
Image Signal Processor
Up to 480 MP/s ISP with HDR encoding for challenging lighting
Video Codec
H.264 and H.265 hardware encode and decode
Camera Interfaces
Two 22-pin Raspberry Pi CSI and two 40-pin MIPI CSI-2 connectors
Camera Support
Up to 8 cameras on SK-AM68, up to 12 cameras on SK-AM69
SK-AM69 Connectivity
Three SuperSpeed USB Type-A, one SuperSpeed USB Type-C, one Ethernet
Expansion Slots
M.2 key E and M.2 key M connectors on SK-AM69
Industrial Interfaces
Four CAN-FD interfaces and four UART channels on SK-AM69
Emulation
On-chip emulation logic for debugging with Code Composer Studio
Software
Processor SDK for Linux and RTOS; TI Deep Learning Library and Edge AI Model Zoo
Framework Support
Edge Impulse integration with TI-optimized model deployment
Automotive Variants
Q1 parts are AEC-Q100 qualified for high-reliability automotive use
SK-AM69 List Price
USD 299 for the AM69 starter kit evaluation module
SK-AM62A List Price
USD 249 for the AM62A-LP starter kit
Target Applications
Machine vision, traffic monitoring, retail automation, factory automation
Overview
Texas Instruments' AM6xA family pairs Arm Cortex-A72 application cores with a hardware deep-learning accelerator, and the SK-AM69, SK-AM68 and SK-AM62A starter kits are the official evaluation modules for that silicon. The AM69A variant reaches up to 32 TOPS across eight Cortex-A72 cores, the AM68A delivers 8 TOPS with two 64-bit Cortex-A72 cores, and the AM62A addresses lower-power vision endpoints.
Each kit is built for multi-camera edge vision. The SK-AM69 exposes two 22-pin Raspberry Pi CSI connectors and two 40-pin MIPI CSI-2 expansion connectors for up to twelve cameras, while the SK-AM68 supports up to eight. The integrated image signal processor handles up to 480 MP/s with HDR encoding, and hardware H.264/H.265 codecs offload video pipelines, so classification and object detection run at low latency and high frame rate.
On the software side the kits are supported by the Processor SDK for Linux and RTOS, the TI Deep Learning Library and the TI Edge AI Model Zoo, with Code Composer Studio for on-chip emulation and debugging. Edge Impulse provides a low-code training and deployment path that targets the same optimized models. Automotive-grade Q1 parts with AEC-Q100 qualification mirror the same architecture. QS Compute supplies the AM6xA starter kits and accessory camera modules.
Key Benefits
Hardware AI acceleration at the edge: 8 TOPS on AM68A up to 32 TOPS on AM69A with a dedicated deep-learning accelerator. Multi-camera scaling: up to 8 or 12 MIPI CSI-2 cameras on a single board. Complete toolchain: Processor SDK Linux/RTOS, TI Deep Learning Library, Edge AI Model Zoo and Code Composer Studio support. Automotive-grade option: AEC-Q100 qualified Q1 parts for high-reliability designs.
Applications
Machine vision inspection systems, traffic monitoring and intelligent transport, retail automation and checkout analytics, factory automation and robotics, autonomous mobile robots, and edge AI box products requiring multi-camera inference.
Request a Quote — TEXAS INSTRUMENTS SK-AM69 AND SK-AM68 — SITARA VISION AI STARTER KITS
QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.
Get Your Quote →