For developers and enthusiasts who create immersive digital worlds, the news is significant: NVIDIA’s powerful physics engines, PhysX and Flow, are now entirely open source. This represents a substantial shift, opening up exciting new possibilities for collaboration and innovation within the developer community.
PhysX has been a familiar name in the realm of realistic simulation for many years, powering everything from dramatic explosions and fluid dynamics in video games to complex physical interactions in various applications. While parts of the PhysX SDK were released under the permissive 3-Clause BSD license over six years ago, a crucial piece remained proprietary: the source code for the high-performance GPU simulation kernel. This was a key element, essential for leveraging the power of graphics processors to handle demanding physics calculations that add realism and depth to digital experiences.
The recent announcement changes that. NVIDIA has now released the source code for these previously closed components, including the vital GPU simulation kernel, making the entire PhysX library fully available on platforms like GitHub. This means developers worldwide now have unprecedented access to the technology’s inner workings.
Adding to the impact, NVIDIA also announced that Flow, another sophisticated simulation engine known for creating visually stunning effects like realistic fire, smoke, and fluid behavior, has joined PhysX in becoming completely open source. The combination of fully open PhysX and Flow puts powerful tools for dynamic visual and physical simulations into the hands of creators everywhere.
This move to complete open source is more than just a licensing change; it’s an invitation to the global development community. It fosters transparency, encourages contributions, and allows developers to customize, optimize, and integrate these physics engines into their projects in ways that weren’t possible before. For independent developers, open-source game engines, and researchers, this provides access to state-of-the-art simulation technology, potentially leading to more realistic and dynamic interactive experiences across a wider range of applications. It’s a step that empowers creators and promises exciting advancements driven by community collaboration.
A major update is here! We are pleased to announce that the latest release of the NVIDIA PhysX SDK now includes the full source code for its GPU acceleration components. This code is being made available under the flexible BSD-3 license, empowering developers with unprecedented access.
The GPU-accelerated features within PhysX are critical for high-performance real-time simulations. Leveraging the thousands of CUDA cores on modern GPUs, these features enable sophisticated physics calculations, including the simulation of rigid body interactions, fluid dynamics, and the behavior of deformable objects. This represents some of the most advanced real-time simulation work built on the CUDA platform.
By open-sourcing this code, our goal is to provide the community with a powerful educational and developmental asset. We encourage developers to explore the implementation details, understand the algorithms behind these simulations, and use this foundation for their own projects and advancements in GPU programming and real-time physics.
Furthermore, the GPU compute shader implementation for our Flow SDK is also being released as open source. Flow is recognized for its capability in real-time fluid simulation, utilizing sparse grids technology for efficient and detailed dynamic fluid effects. Opening this component allows deeper insight into advanced compute shader programming for simulation tasks.
We look forward to the contributions and innovations that the community will bring using these newly open resources. Empowering developers with access to this source code is a key step in fostering broader understanding and accelerating the evolution of real-time simulation technology.
Choose the option that best fits the desired tone and specific aspects you want to emphasize. Both are around the requested length and adhere to your requirements.
Orange Man 
Klaudia