//Talks
Upcoming talks.
What governs inference speed on hardware people actually own, and how to measure it yourself. Listed only once a session is accepted or booked.
- Accepted
- 1
- Booked
- 1
- Topic
- Edge inference
On the calendar.
-
Feb 2027
Embedded Vision Summit 2027
Talk, accepted
Accepted September 2026: "What actually governs vision inference on a Raspberry Pi 5", a 20-minute session at the 2027 Embedded Vision Summit, San Francisco, 2 to 4 February 2027. The curated program reviews speaker slides in three draft rounds before the event.
-
Dec 2026
Austin Python Meetup
Talk, booked
Booked for Wednesday 9 December 2026, 7:00 PM, a hybrid evening event in downtown Austin. A hands-on session with llama-roofline: opening real GGUF model files to see what they contain, then measuring what local LLM token generation costs on an ordinary laptop. Bring a laptop to run it yourself.
Want to try it before the talk?
The tools the talks use are on PyPI and run on a laptop with no configuration.
Install commands, documentation and DOIs are on the tools page. To ask about a talk, email manunicholasjacob@gmail.com.