Work

Guardian: Safe GPU Sharing in Multi-Tenant Environments

Year: 2024

Type: conference-paper

Abstract: Modern GPU applications, such as machine learning (ML), can only partially utilize GPUs, leading to GPU underutilization in cloud environments. Sharing GPUs across multiple applications from different tenants can improve resource utilization and consequently cost, energy, and power efficiency. However, GPU sharing creates memory safety concerns because kernels must share a single GPU address space. Existing spatial-sharing mechanisms either lack fault isolation for memory accesses or require static partitioning, which leads to limited deployability or low utilization.

Source: International Middleware Conference

DOI: 10.1145/3652892.3700768

Authors: Manos Pavlidakis, Giorgos Vasiliadis, Stelios Mavridis, Anargyros Argyros, Antony Chazapis, Angelos Bilas

Institutions: FORTH Institute of Computer Science, Foundation for Research and Technology Hellas, Hellenic Mediterranean University, University of Crete

Language: English

Topic: Parallel Computing and Optimization Techniques

Subfield: Hardware and Architecture

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

Funders: European High Performance Computing Joint Undertaking, European Commission

Awards: Pilot using Independent Local & Open Technologies, G434337789