GPU-Virt-Bench: A Comprehensive Benchmarking Framework for Software-Based GPU Virtualization Systems

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: VG, Jithin, PS, Ditto
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917171445104640
author VG, Jithin
PS, Ditto
author_facet VG, Jithin
PS, Ditto
contents The proliferation of GPU-accelerated workloads, particularly in artificial intelligence and large language model (LLM) inference, has created unprecedented demand for efficient GPU resource sharing in cloud and container environments. While NVIDIA's Multi-Instance GPU (MIG) technology provides hardware-level isolation, its availability is limited to high-end datacenter GPUs. Software-based virtualization solutions such as HAMi-core and BUD-FCSP offer alternatives for broader GPU families but lack standardized evaluation methodologies. We present GPU-Virt-Bench, a comprehensive benchmarking framework that evaluates GPU virtualization systems across 56 performance metrics organized into 10 categories. Our framework measures overhead, isolation quality, LLM-specific performance, memory bandwidth, cache behavior, PCIe throughput, multi-GPU communication, scheduling efficiency, memory fragmentation, and error recovery. GPU-Virt-Bench enables systematic comparison between software virtualization approaches and ideal MIG behavior, providing actionable insights for practitioners deploying GPU resources in multi-tenant environments. We demonstrate the framework's utility through evaluation of HAMi-core, BUD-FCSP, and simulated MIG baselines, revealing performance characteristics critical for production deployment decisions.
format Preprint
id arxiv_https___arxiv_org_abs_2512_22125
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GPU-Virt-Bench: A Comprehensive Benchmarking Framework for Software-Based GPU Virtualization Systems
VG, Jithin
PS, Ditto
Distributed, Parallel, and Cluster Computing
Artificial Intelligence
The proliferation of GPU-accelerated workloads, particularly in artificial intelligence and large language model (LLM) inference, has created unprecedented demand for efficient GPU resource sharing in cloud and container environments. While NVIDIA's Multi-Instance GPU (MIG) technology provides hardware-level isolation, its availability is limited to high-end datacenter GPUs. Software-based virtualization solutions such as HAMi-core and BUD-FCSP offer alternatives for broader GPU families but lack standardized evaluation methodologies. We present GPU-Virt-Bench, a comprehensive benchmarking framework that evaluates GPU virtualization systems across 56 performance metrics organized into 10 categories. Our framework measures overhead, isolation quality, LLM-specific performance, memory bandwidth, cache behavior, PCIe throughput, multi-GPU communication, scheduling efficiency, memory fragmentation, and error recovery. GPU-Virt-Bench enables systematic comparison between software virtualization approaches and ideal MIG behavior, providing actionable insights for practitioners deploying GPU resources in multi-tenant environments. We demonstrate the framework's utility through evaluation of HAMi-core, BUD-FCSP, and simulated MIG baselines, revealing performance characteristics critical for production deployment decisions.
title GPU-Virt-Bench: A Comprehensive Benchmarking Framework for Software-Based GPU Virtualization Systems
topic Distributed, Parallel, and Cluster Computing
Artificial Intelligence
url https://arxiv.org/abs/2512.22125