Accelerating Software on the Edge (EMEA)
The rise of edge computing has increased the need for portable hardware acceleration for AI and image processing, especially where power consumption matters or cloud connectivity is unreliable.
Automotive drives software innovation in vehicles, particularly for driver assist and self-driving features, motivating neural networks like Llama and GPT. Traditional algorithms like BLAS and FFT also benefit from running faster on GPU-like processors than on CPUs.
However, challenges like power consumption, cost, and safety-critical requirements remain. Developers must develop on cloud or desktop infrastructure and migrate to final hardware, increasing the demand for software portability.
This session brings together experts from the semiconductor and software industries to explore new hardware challenges and opportunities for edge computing and the crucial role of open standards. The UXL Foundation and Autoware collaboration showcases this potential.
