3 Repos
Container runtime wrappers that automatically expose host graphics processing units to isolated environments.
Distinct from Container Runtime Wrappers: Distinct from Container Runtime Wrappers: focuses on GPU-specific hardware exposure rather than general engine management.
Explore 3 awesome GitHub repositories matching devops & infrastructure · GPU Runtime Wrappers. Refine with filters or upvote what's useful.
NVIDIA Docker is a container runtime wrapper that enables the use of host-level graphics processing units within isolated container environments. It functions as a containerized GPU orchestrator, mapping physical hardware resources into virtualized environments to support high-performance computing and machine learning workloads. The project provides a toolkit that facilitates integration between containerized applications and host-level graphics hardware. By utilizing a pre-start hook to intercept container creation, the runtime injects necessary device drivers and libraries into the isolate
Wraps container runtimes to automatically expose host graphics processing units for hardware-accelerated tasks.
GameMode is a Linux system daemon that automatically applies CPU, GPU, and I/O performance optimizations when a game starts and reverts them when the game exits. It coordinates system-level tweaks by reading a plain-text configuration file, switching the CPU frequency scaling governor to performance mode, adjusting the I/O scheduler for low latency, and setting the graphics card to a high-performance power state through direct sysfs or driver interface writes. The project includes a daemon that communicates with game launchers via Unix domain sockets, a hybrid GPU launcher that wraps game exe
Prepends a GPU selection command to game launches to ensure optimizations target the correct graphics adapter.
Das NVIDIA GPU Container Toolkit ist eine Reihe von Tools, die entwickelt wurden, um NVIDIA GPU-Beschleunigung innerhalb von OCI-konformen Containern für Rechen- und Grafik-Workloads zu ermöglichen. Es fungiert als OCI-Container-Runtime-Wrapper und Linux-GPU-Treiber-Brücke, die Host-NVIDIA-GPU-Treiber in den Container-Namespace mappt, um direkten Hardwarezugriff bereitzustellen. Das Toolkit enthält einen GPU-beschleunigten Container-Builder zur Erstellung von Images, die vorkonfiguriert sind, um Hardwarebeschleunigung zu nutzen. Dies ermöglicht die Integration von Grafikprozessoren in die Container-Umgebung sowohl für das Erstellen als auch für das Ausführen rechenintensiver Workloads. Das Projekt deckt eine Reihe von High-Performance-Computing-Domänen ab, einschließlich Workflows für das Training von KI-Modellen, GPU-basiertes Cloud-Computing und wissenschaftliche Simulationen.
Functions as an OCI container runtime wrapper that automatically exposes host GPUs to isolated environments.