How to turn your ideas in AI into products
summary
Distributor shows how to move pre-trained models from cloud to edge using supplier toolchains and hardware: model conversion with X-Cube.AI and pruning, deployment onto Cortex-M microcontrollers (STM32 M0–M7, 3–400 MHz) for low‑power inference, or onto FPGA acceleration cards (Alveo U200/U250/U280, upcoming 7 nm Versal with HBM) for high‑throughput workloads; NXP parts scale from M4 to 16x A53 for gateways; claims production workflows, logistics and part sourcing in place and edge inference workflows already working with cloud options (AWS F1).
AI-written summary · about
chapters (6)
- 0:00Introduction and company overview
- 3:44AI applications across industries
- 6:37AI in automotive and edge processing
- 10:03FPGA solutions for AI acceleration
- 13:23Software-defined acceleration and development tools
- 15:59Choosing the right AI hardware
related talks
04mMultiple embedded and Cloud AI solutions at ST
24mRunning AI/Neural networks on microcontrollers made simple with the STM32Cube.AI
42mThe Stakes of Green AI for Embedded Systems
20mArchitecting AI: Efficient Models, Everywhere
24mInnovation and Solutions for Embedded and Edge AI
50mAI acceleration at the edge
19mFrom concept to production: design and optimize computer vision applications on Arm
28mML on the Edge for Industry 4.0 with Arm Ethos N-78 Neural Processing Unit