Leading the development of a distributed computing framework that enables GPU resource sharing among edge servers to run computer vision, digital twins, and other AI applications. The open-source GVirtuS has been updated to run CNNs transparently on remote GPUs, and we are currently upgrading GVirtuS to support PyTorch for running YOLO and LLMs. This work is part of the European project CLEVER, collaborating with 20 academic and industrial partners on embedded AI solutions.
[ad_2]
source
