But can it run AI?

NPU, TPU, MPU, GPU, MCU: feels like it is already one three-letter abbreviation too many! In today's video I briefly explain the main types of compute hardware and how well they are suited for AI and machine learning. From general-purpose processors to specialized accelerators: their strengths, limitations, and the common gotchas. Boards shown in the video: Raspberry Pi AI HAT+, Build-in Hailo-8 AI Accelerator https://s.click.aliexpress.com/e/_EG4CF8o Raspberry Pi 5 BCM2712 processor https://s.click.aliexpress.com/e/_ExScjyg Raspberry Pi Pico Board RP2040 https://s.click.aliexpress.com/e/_ExolQDq Raspberry Pi Pico 2 RP2350 https://s.click.aliexpress.com/e/_EwEt8Fa Seeed Studio XIAO ESP32-S3 https://s.click.aliexpress.com/e/_EHh81Mu Resources used: https://www.youtube.com/watch?v=BML1YHZpx2o https://www.youtube.com/watch?v=FTrQNmI3G0s https://www.youtube.com/watch?v=KVedp1Ij--g