Practical experiences in building a low-latency real-time Object Detection pipeline, from handling initial blocking issues to a stable streaming architecture.
Explore all notes
Filter by topic and browse all posts in reverse chronological order.
An examination of how data pipeline design impacts PyTorch training performance, supported by simple experiments and benchmarks.
A deep-dive journey inside the ESP-WROOM-32 chip, exploring the Xtensa CPU, memory architecture, and how to reverse-engineer firmware to uncover its secrets.
A real-world story of connecting a camera and a PC via Ethernet, exploring concepts from PoE and static IP to turning a laptop into a DHCP server.
Dive deeper into CUDA to uncover the principles and practices behind high-performance GPU computing.
Learn about GPU architecture and the CUDA model to understand how parallel hardware operates and optimize performance.
Page 1 / 2
PreviousNext