multi-node

Rust hosting control panel for Debian. Multi-node clustering, kernel-enforced quotas, tamper-evident audit log, and live progress bars on every operation. The Cloudpanel / HestiaCP alternative that doesn't shell-template PHP at root.

Independent full-source DrNetwork multi-node management panel based on S-UI

Terraform provider for 3x-ui (Xray/VLESS/Reality/WireGuard) — manage inbounds, clients, panel settings, cluster nodes & Xray config as code. GitOps for your VPN panel.

Best Multi-Device AI Dev Console 2026: Manage Claude & Codex Sessions Remotely

Java Client SDK for Stateful.co

A comprehensive, self-hosted web panel for managing OpenVPN, Cisco AnyConnect, L2TP/IPsec, WireGuard, and Sing-box servers, along with users, resellers (sub-admins), and multi-node deployments — featuring a rich UI and full API support.

Enterprise Multi-Node Hybrid RAG, PyTorch Neural Vector Reranker & GPU Compute Console

Super Ollama Load Balancer - Performance-aware routing for distributed Ollama deployments with Ray, Dask, and adaptive metrics

一个浏览器继续所有电脑上的 DeepSeek Harness 会话：自托管、多节点、仅出站 WSS、官方风格 Web UI。

Local distributed-development harness and proving ground for the full stack: repeatable single-node and multi-node boot, Postgres-backed durability drills, fault injection, trace plumbing, end-to-end examples, and operator-facing smoke paths for outer_brain, Citadel, jido_integration, and execution_plane development.

Turn spare hardware into a production homelab. Ansible playbooks for Docker Swarm, Traefik, and 30+ services with automated setup and DNS.

RadPanel — a modern, minimal web control panel for Xray-core. One-command install, clean multi-language UI, multi-protocol & multi-node. A friendly GPL-3.0 fork of 3x-ui.

Synclave — encrypted multi-node backup + replication + agent mesh (A2A). Ed25519+X25519 identity, conversation ledger, GPU offload, 20-node scale. Formerly hermes-sync.

DeepSeek-V4.1-Flash (dealignai UNCENSORED-FP8) serving at 1M context on 8x NVIDIA DGX Spark GB10, vLLM TP8 + DSpark: tonyd2wild's four-node recipe carried to eight ranks, with measurements. Serving since 2026-09-11.

Run a model too big for one GPU in LM Studio across two PCs over LAN (distributed inference via llama.cpp RPC): two DLLs + one env var GGML_RPC_SERVERS, no LM Studio changes. Windows/Vulkan/AMD tested. Home-lab hack by Robert + Kod (AI).

The LUMI AI Guide is designed to assist users in migrating their machine learning applications from smaller-scale computing environments to the LUMI supercomputer.

GLM-5.3 (uncensored) on 8x NVIDIA DGX Spark GB10 with vLLM TP=8: node launchers, measurement scripts, ops helpers, and the config-with-evidence behind the HF field report

Free Docker-free multi-node game and application hosting control plane written in Rust

Measured vLLM serving for 1–2 NVIDIA DGX Spark (GB10): TP=2 RoCE flagship, validation ledger, stranger-ready Path A/B onboarding.

Build NCCL-Tests and configure SSHD in PyTorch container to help you test NCCL faster!