Overview
========

.. contents::
   :depth: 2
   :backlinks: none

What is Genv
------------
.. note::

   You can jump directly to the quick start toturial :doc:`here <quickstart>`.

Genv is an open-source project that brings environment and cluster management capabilities to GPUs.

The project started as an environment management software for GPUs (hence the name - GPU environments), and was highly inspired by projects like `PyEnv <https://github.com/pyenv/pyenv>`_ and `Conda <https://docs.conda.io/projects/conda/en/latest/>`_.
If you are not familiar with the concept of environments, you can check out the documentation `venv <https://docs.python.org/3/library/venv.html>`__, `pyenv <https://github.com/pyenv/pyenv>`__ and `Conda <https://docs.conda.io/projects/conda/en/stable/>`__ as a reference.

Over time, the project developed into a complete GPU cluster management tool with various capabilities and mechanisms like :doc:`monitoring <../usage/monitoring>`, :doc:`enforcement <../usage/enforcement>`, :doc:`remote capabilities <../remote/overview>` and more.

In addition to running GPU environments in a cluster, Genv also supports managing :doc:`LLMs <../llms/overview>` on GPUs.

At its core, Genv is a `Python package <https://pypi.org/project/genv>`__ that provisions GPUs to environments and keeps its state as :ref:`files <Files>` at :code:`/var/tmp/genv`.

On top of this core layer, Genv has a CLI (i.e. the command :code:`genv`), :doc:`Python SDK <../usage/python-sdk>` and :ref:`integrations <Integrations>` with many common tools and environments such as terminal, containers (e.g. :code:`docker`), Ray, Visual Studio Code, JupyterLab and PyCharm.

Genv is built with a decentralized architecture, meaning that different :code:`genv` commands run separately and coordinate with one another and that there is no centralized service that handles the requests.

.. figure:: overview.png

   GPU environments with Genv

Architectures
-------------
Genv supports two main architectures: local for a single machine and :doc:`remote <../remote/overview>` for a multi-machine cluster.

Local
~~~~~
.. figure:: architecture-local.png

   Genv local architecture

When running in local mode, Genv manages the GPUs on the local machine where Genv is running.
This is mostly relevant for the case of a single user with GPUs in his or her machine, or the case of multiple users that share a single GPU machine.

Remote
~~~~~~
.. figure:: architecture-remote.png

   Genv remote architecture

When running in remote mode, Genv manages the GPUs on any number of remote machines over SSH.
This is mostly relevant for the case when multiple users that share multiple GPU machines.

For more information, see :doc:`here <../remote/overview>`.

Where to Go Next
----------------
If you are not familiar with Genv, it is recommended to follow the :doc:`quick start tutorial <quickstart>`.
If you are into LLMs, check out the :doc:`LLMs overview <../llms/overview>`.

If you are looking for how to install Genv, visit the :doc:`installation page <installation>`, and if you want to know how to do specific things, check out the :doc:`usage guide <../usage/usage>`.