https://github.com/fabioz/optimuspy

Python Launcher with Optimus Enablement
https://github.com/fabioz/optimuspy

Last synced: about 17 hours ago
JSON representation

Python Launcher with Optimus Enablement

• Host: GitHub
• URL: https://github.com/fabioz/optimuspy
• Owner: fabioz
• License: mit
• Created: 2017-08-22T10:40:54.000Z (about 7 years ago)
• Default Branch: master
• Last Pushed: 2017-08-19T12:00:58.000Z (about 7 years ago)
• Last Synced: 2024-10-20T06:12:25.661Z (24 days ago)
• Language: Python
• Size: 4.88 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# optimuspy

Python Launcher with Optimus Enablement and AMD equivalent

## Problem & Solution

I am using python for developing 3D applications and games. I have a dedicated GPU and an Intel HD Graphics 630. My applications run with the Intel HD Graphics 630 unless changed in the nVidia control panel.

I started googling then I found a solution how to programatically enable the dedicated graphics card. This problem only occurres on Optimus Systems. The solution is simple. An integer variable called _NvOptimusEnablement_ should be exported in order to enable the nVidia card.

Unfortunately exporting this variable from a python module **does not** give access to the dedicated graphics card. Some of the python modules are DLLs called PYD files. Importing pyd file with an exported _NvOptimusEnablement_ is not working properly.

The only solution I found so far is to create a python launcher that exports _NvOptimusEnablement_. From now on I am able to test my applications and games using the integrated and the dedicated graphics card as well.

Reference: [Optimus Rendering Policies](http://developer.download.nvidia.com/devzone/devcenter/gamegraphics/files/OptimusRenderingPolicies.pdf)

## How does it work?

The entire source code is in the [optimuspy.cpp](optimuspy.cpp).

The program simply passes every argument to the Py_Main.

To create a launcher for a specific application you can hardcode the parameters and add an icon using a resource file.

## Install

The [build.py](build.py) will detect the python installation and use the proper compiler arguments to create the optimuspy launcher.

Using the `--install` the `optimuspy.exe` will be placed next to the `python.exe`.

```batch

git clone https://github.com/cprogrammer1994/optimuspy

cd optimuspy

python build.py --install

```

### Compile using gcc

```batch

python build.py --gcc --install

```

## Sample

```batch

> python -m ModernGL

ModernGL: 4.1.11

Vendor: Intel

Renderer: Intel(R) HD Graphics 630

Version: 4.4.0 - Build 21.20.16.4541

```

```batch

> optimuspy -m ModernGL

ModernGL: 4.1.11

Vendor: NVIDIA Corporation

Renderer: GeForce GTX 1050/PCIe/SSE2

Version: 4.5.0 NVIDIA 384.94

```