Simplifying & Scaling Edge AI

on YouTube ↗15m

" summary

Claim: Edge AI runs today on MCU and MPU hardware using Zephyr and Linux with EIQ suite toolchains and ready models. Demonstrated stack uses NXP i.MX93 and MCX N with an NPU that is 172× faster on NPU than MCU for the same power, a Tria MaxSport board qualified for production, and cloud integration via /IOTCONNECT powered by AWS sending aggregated alerts after on‑device face, landmark and iris models run in parallel.

AI-written summary · about

" chapters (7)

1. 0:00Edge AI collaboration
2. 2:12NXP edge AI stack
3. 5:22Modular AI with EIQ
4. 6:04Live demo: worker safety
5. 9:20Scaling edge AI with IoT Connect
6. 10:23AWS on AI at the edge
7. 13:16Summary and call to action

" related talks

• 05mEdge AI: Because the Cloud is Too Slow...Said Ouissal · Kubernetes on Edge Day Europe 2026
• 18mThe Cutting Edge of AI and IoTIvan Turasov · The Things Conference 2024
• 26mAI computing power in Edge Industry use casesEvgeniy Vlasov · Rise of AI Summit 2020
• 19mDistributed AI for Automated and Scalable Quality Inspection using EDGE Computing...Igor Khapov · OpenPOWER Summit NA 2020
• 33mAI on the Edge – Unlocking New Opportunities in IoTBenjamin Cabé, Pamela Cortez · Open Source Summit + Embedded Linux Conference North America 2019
• 44mArtificial Intelligence in Small Embedded SystemsDov Nimratz · Embedded Fest 2019
• 01h 19mWORKSHOP 2: AI FOR EVERYONESean Dong · droidcon Berlin 2018
• 14mKeynote: Empowering Innovation in Embedded Systems: Integrating AI, IoT and Edge Computing for Smarter SolutionsPatrick Johnson · RISC-V Summit North America 2024