Giving developers the keys to building the future
summary
Claimed hardware and software details: Cortex‑M Helium extensions yield Cortex‑M Helium: 15x ML uplift for tiny‑ML; TF‑Lite‑Micro and Arm’s CMSIS‑NN give 4–5x improvements and 10x model‑size cuts via quantization/pruning/compression. Concrete demos: AquaSecor moves leak detection from cloud to on‑device ML (battery life jumps from weeks to years), Audio Analytic runs a baby‑monitor on Cortex‑M0+ with fixed‑point tuning, Shoreline IoT ships Cortex‑M4 devices with TF‑Lite‑Micro for five‑year battery life. Cloud/edge stack: Pelion data management ingesting 2–3M events/sec plus edge SDKs, and a UNICEF Malawi data‑science challenge with open flood‑prediction datasets.
AI-written summary · about
chapters (20)
- 0:00Introduction
- 5:28Hardware vs software
- 7:25IOT
- 10:35A F3
- 13:17AI
- 14:19Aqua Secor
- 16:42Audio Analytics
- 17:58Industry Insights
- 20:36Data Consumption
- 21:53New wave of computing
- 23:41Jen
- 33:17Shoreline IoT
- 34:41ARM Ecosystem
- 36:34Ecosystem
- 38:07John
- 41:09Cloud
- 45:07Devices Embedded ML
- 47:59Recap
- 52:05What happens when you apply technology
- 53:15UNICEF Innovation Lab
related talks
46mSoftware Development Tools for IoT
25mScaling IoT in Practise: From PoCs to Deployment
47mIntroducing the Next Generation of Windows on Arm
25mChallenging Hardware Contests With Zephyr and Lessons Learned
13mOpenHW CORE-V MCU DevKit for Cloud Connected IoT
17mSay Goodbye to Fear, Uncertainty, and Doubt: Innovate with Codasip Studio Fusion
27mInsights on the Re-Architecting to a Single Codebase, Running on Different Hardware...
49mOne Code Base, Multiple Targets