backend.ai

Backend.AI is a streamlined, container-based computing cluster platform that hosts popular computing/ML frameworks and diverse programming languages, with plugg

What is backend.ai?

Backend.AI is a streamlined, container-based computing cluster platform that hosts popular computing/ML frameworks and diverse programming languages, with pluggable heterogeneous accelerator support including CUDA GPU, ROCm GPU, Gaudi NPU, Google TPU, GraphCore IPU and other NPUs.

Overview of backend.ai

What it is

Backend.AI is a streamlined, container-based computing cluster platform that hosts popular computing/ML frameworks and diverse programming languages, with pluggable heterogeneous accelerator support including CUDA GPU, ROCm GPU, Rebellions, FuriosaAI, HyperAccel, Intel Gaudi, Tenstorrent, Google TPU, Graphcore IPU and other NPUs. It allocates and isolates the underlying computing resources for multi-tenant computation sessions on-demand or in batches with customizable job schedulers using its own orchestrator named Sokovan. All functions are exposed as REST and GraphQL APIs.

Key features

  • Container-based computing cluster for multi-tenant sessions
  • Pluggable heterogeneous accelerator support (CUDA GPU, ROCm GPU, Intel Gaudi, Google TPU, and others)
  • Customizable job schedulers with the Sokovan orchestrator
  • REST and GraphQL API exposure
  • Web-based access to compute sessions via Jupyter, web terminal, SSH, and VSCode
  • Virtual folder (vfolder) abstraction for network-based storage
  • App proxy for secure access to in-container applications

Use cases

Backend.AI is suitable for research and production environments requiring multi-tenant access to heterogeneous computing resources. Common use cases include running machine learning training jobs, hosting Jupyter notebooks for data science teams, providing on-demand GPU/TPU sessions for experiment reproducibility, and managing batch job queues for clusters with diverse accelerators.

Getting started

Quick start for development:bashgit clone https://github.com/lablup/backend.ai.gitcd backend.ai./scripts/install-dev.shThen start each component in separate terminals: manager, agent, storage proxy, web server, account manager (optional), and app proxy (optional). Access the Web UI at http://localhost:8090.For production or multi-node setups, consult the documentation or contact the sales team at [email protected] for professional support and deployment options.

License and pricing

License: LGPL-3.0. Pricing: open source.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/lablupbackendai.svg)](https://yunodb.com/tools/lablupbackendai?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/lablupbackendai?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/lablupbackendai.svg" alt="Featured on YunoDB" height="20"></a>