Viviane Potocnik from ETH Zürich shares how her team is tackling the challenges of Edge computing within the dAIEDGE project. The ETH Zurich team is dedicated to designing hardware interfaces for advanced sensors that work with event- and frame-based data. These technologies enable faster and more efficient processing, optimising performance on Edge devices where resources are limited. ‘Our involvement in dAIEDGE is focused on delivering innovative solutions that push the technology to the next level,’ explains Potocnik.
The proposed dAIEDGE ecosystem seeks to address a critical challenge: bringing advanced models, such as large language models (LLMs), to small devices. ‘Although LLMs are evolving rapidly, we don't have precise solutions to adapt them to edge devices,’ says Potocnik. This approach is especially relevant in the current context, where energy efficiency and low-power solutions are essential.
Potocnik also underlines that initiatives such as dAIEDGE can inspire both companies and research institutions to delve deeper into the challenges of edge computing, fostering collaboration and innovation to build a more sustainable future.
Listen to the full interview with Viviane Potocnik here: https://www.youtube.com/watch?v=iSKHpWwPTqM&ab_channel=dAIEDGE