GPUPool: A Holistic Approach to Fine-Grained GPU Sharing in the Cloud
Published in International Conference on Parallel Architectures and Compilation Techniques (PACT), 2022
GPUPool proposes a holistic approach to fine-grained GPU sharing in cloud environments. It addresses the challenge of low GPU utilization in cloud data centers by enabling multiple workloads to efficiently share GPU resources, improving overall resource efficiency and reducing costs.
Recommended citation: Xiaodan Serina Tan, Pavel Golikov, Nandita Vijaykumar, Gennady Pekhimenko. (2022). "GPUPool: A Holistic Approach to Fine-Grained GPU Sharing in the Cloud." PACT 2022. pp. 317-332.
Download Paper