Figures
All diagrams and plots on this site are original, drawn as SVG for this course; the plotted data are either synthetic (and labelled as such), computed from public datasets (the scikit-learn digits dataset), or order-of-magnitude teaching values with the sources of their underlying numbers named in the caption. The SVG files are in assets/fig/ and may be reused for teaching with attribution.
Photographs
Photographs are taken from Wikimedia Commons under the licences listed below and are stored in assets/img/. Each is credited beneath its use. A photograph that is missing from the folder is simply not shown.
| File | Shows | Source and author | Licence |
|---|---|---|---|
nucleo-board.jpg | STM32 Nucleo microcontroller development board | not downloaded yet | — |
esp32-module.jpg | ESP32 Wi-Fi/Bluetooth microcontroller module | not downloaded yet | — |
arduino-nano.jpg | Arduino Nano format board | not downloaded yet | — |
raspberry-pi-4.jpg | Raspberry Pi 4 Model B single-board computer | not downloaded yet | — |
mems-accelerometer.jpg | MEMS 3-axis accelerometer breakout board | not downloaded yet | — |
mpu6050.jpg | MPU-6050 6-axis IMU module | not downloaded yet | — |
electret-microphone.svg | Electret condenser microphone schematic | not downloaded yet | — |
stm32-chip.jpg | STM32 Cortex-M microcontroller package | not downloaded yet | — |
bcm2711.jpg | Broadcom BCM2711 SoC of the Raspberry Pi 4 | not downloaded yet | — |
jetson-nano.jpg | NVIDIA Jetson Nano developer kit | not downloaded yet | — |
gpu-die.jpg | GPU die photograph | not downloaded yet | — |
fpga-die.jpg | FPGA die photograph | not downloaded yet | — |
fpga-board.jpg | Small FPGA development board | not downloaded yet | — |
Fonts
IBM Plex Sans Condensed and IBM Plex Mono (© IBM Corp.) and Source Serif 4 (© Adobe), all under the SIL Open Font License 1.1, self-hosted from assets/fonts/ so that no third-party font service is contacted. Licence texts are included in that folder.
Privacy
The site uses no cookies, analytics or external requests. The only thing stored in your browser is your light/dark theme choice, in local storage.
Sources and acknowledgements
The selection of topics was informed by comparable courses: MIT 6.5940 TinyML and Efficient Deep Learning Computing, Harvard CS249r Tiny Machine Learning and its open textbook Machine Learning Systems, and ETH Zürich's Machine Learning on Microcontrollers. Quoted numerical results are attributed to their papers in the text; see the references.