Open Source Hardware: Hacking Silicon for Fun (instead of profit)
summary
Argues that the open silicon stack now makes small-scale chip fabrication practical, claiming less than $700,000 to start building useful parts, showing progress with tools like yosis and layout flows (Kad, Magic) and prototypes such as the pulpo chip and Beagle V-based cores. Calls for a 500 nm laser direct-write Fab on a farm, an open InfiniBand ASIC and boards with 16 GB of RAM (2 GB is too small) so students can run real workloads and tape out reproducible chips.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:30From farming to open source silicon
- 3:00Silicon self-replication myth
- 4:27Can farmers build chips?
- 5:32Beagle V Fire progress
- 6:39Linux-level revolution
- 7:44Open source hardware tools
- 9:10Open source hardware in space
- 10:14Human-AI systems
- 11:40Open source GPU training
- 12:45Grad student platforms
- 13:29Closing
related talks
16mOpen-source silicon chips
43mOpen Source Silicon: Where software ends and hardware begins
18mOpen source chip design: for fun and for profit
01h 05mChallenges and opportunities for Open Source Hardware in academia and industry: Roundtable
26mFrom Tapeouts to Products: Why Open-Source Silicon Must Solve Real Problems to Survive
15mBuilding an Open Source Hardware Ecosystem: From Foundations to Rooftops
40mFree and Open Source Software Tools for Making Open Source Hardware
40mOpen-Source HW Commercial Adoption: Lessons Learned