# Tanawin Devaveja

- Institution: CERN
- Department: Summer Student Lecture Programme General
- Speaker: Tanawin Devaveja
- Year: 2025
- Language: English
- Source: [CERN](https://videos.cern.ch/record/3023212)
- Licence: [CERN](http://copyright.web.cern.ch)
- Who can watch: Teens profiles and older (sources give no ratings; placed by Invisible Network from the source's own words)

**DFX4ML: enable Dynamic Function eXchange for Machine Learning** In resource-constrained edge computing applications like satellites, autonomous vehicles, and wearable devices, power efficiency and hardware limitations are critical. FPGAs offer reconfigurability, but large ML models often exceed their capacity. This project tackles the issue by enabling dynamic Function eXecution (DFX), allowing efficient use of FPGA resources.

I'm a master's student at Chulalongkorn University, currently interning at edgespAIce@CERN. My thesis project, KATHRYN, is a hybrid hardware design and simulation framework that aligns closely with my CERN work, both aiming to generate flexible and efficient hardware..

0:00:00 Slide 1

0:05:21 Slide 2

0:06:30 Slide 3

## Lectures

1. [Tanawin Devaveja](https://invisiblenetwork.ai/learn/cern/3023212/4f1945ad4c04), 9 min (CERN's player)