gpu-cluster

Resource scheduling and cluster management for AI

Best practices & guides on how to write distributed pytorch training code

Mimir — 一站式科研工作台插件：LaTeX 论文边写边编译、arXiv 文献管理、实验追踪、指标图表、GPU 服务器 SSH 任务编排，管理科研全周期。An open-source research workbench plugin for the whole research cycle.

A declarative workflow orchestrator for massive GPU clusters of any kind, including slurm, K8S and docker. Topology-aware node and GPU placement, multi-node replicas and sweeps, readiness/failure probes, and batch submission — so one descriptor drives hundreds of GPUs instead of a pile of hand-written bash.

Mixed-vendor GPU inference cluster manager with speculative decoding

AI动力(AI Power) GPU云平台使用指南

qihoo360 xlearning with GPU support; AI on Hadoop

GPU management System on Kubernetes For AI, Deep-Learning, Machine-Learning Researcher.

Example code for deploying GPU workloads on ECS

Complete setup guide for a 2-node NVIDIA DGX Spark cluster — distributed training, CUDA inference with EXO, NCCL tuning for Grace Blackwell, NVMe-TCP shared storage, and 200 Gb/s direct fabric networking.

GPU operator agent for the Scalattice inference network (Rust).

Turn a DGX Spark GPU cluster into a self-running team of specialized AI agents — cluster control, role-specialized workers, and an autonomous review gate for Hermes

Project "Springfield"

Official artifact for MoSim, a trace-driven GPU-cluster simulator with dynamic network contention modeling (IEEE MASCOTS 2026).

Detailed Analysis Traces for AI jobs leveraging spot GPU resources

Audit GPU cluster communication schedules from NCCL logs. Zero dependencies. CI-ready.

Web UI for orchestrating distributed llama.cpp RPC GPU clusters with auto node discovery, telemetry, and one-click deployment.

⚡ Self-registering node agent for distributed AI clusters, dynamically serving local LLMs via vLLM across heterogeneous GPUs.

A reproducible comparison of Kubernetes GPU schedulers (Kueue, Volcano, KAI) on identical workload traces — kwok-simulated GPU fleet, runs on a laptop.