no code implementations • 2 May 2023 • Manos Pavlidakis, Stelios Mavridis, Antony Chazapis, Giorgos Vasiliadis, Angelos Bilas
Today, using multiple heterogeneous accelerators efficiently from applications and high-level frameworks, such as TensorFlow and Caffe, poses significant challenges in three respects: (a) sharing accelerators, (b) allocating available resources elastically during application execution, and (c) reducing the required programming effort.