NVIDIA PhysX and Flow Become Completely Open Source Solutions

NVIDIA PhysX SDK Goes Fully Open Source

Approximately six and a half years after the NVIDIA PhysX SDK transitioned to an open-source model under the 3-Clause BSD license, a significant milestone has been achieved. The latest update, which was recently released on GitHub, now includes the complete source code for the GPU simulation kernel, effectively enhancing the SDK’s openness.

NVIDIA’s Announcement

In their announcement, NVIDIA expressed enthusiasm about this update, stating:

We’re excited to share that the latest update to the PhysX SDK now includes all the GPU source code, fully licensed under BSD-3!

This comprehensive release features over 500 CUDA kernels that facilitate advanced functionalities, including rigid body dynamics, fluid simulation, and support for deformable objects. GPU PhysX exemplifies one of the most sophisticated applications of CUDA and GPU programming in real-time simulation, ultimately serving as a valuable resource for developers and learners alike.

Introduction of the Flow SDK

In conjunction with the GPU simulation kernel source code, NVIDIA is also making available the complete GPU compute shader implementation of the Flow SDK. This library focuses on real-time, sparse grid-based fluid simulation, broadening the scope of open-source resources available to the developer community.

Impact on Users and Gaming

While NVIDIA PhysX may not have been in the spotlight recently, concerns arose when NVIDIA halted support for newer GeForce RTX 50 Series graphics cards. This led users to explore unconventional solutions, such as pairing an RTX 50 GPU with an RTX 3050 dedicated exclusively to handling PhysX tasks.

The implications of this change mainly impact a select few older PC titles like Batman: Arkham Asylum and Mirror’s Edge. However, the recent release of the GPU simulation kernel source code empowers modders to potentially rectify these issues themselves. Although proactive action from NVIDIA would have been preferable, this open-source strategy offers the community an avenue for solutions. We anticipate monitoring upcoming mod releases that will leverage the newly accessible GPU simulation kernel source code.

Source&Images