Borderless Gaming update brings DLSS 5 Frame Generation to unsupported GPUs like the RTX 3090

PC gamers can now use DLSS 5 Frame Generation on graphics cards that NVIDIA does not officially support. This update to the Borderless Gaming application removes the need for complex DLL injections or game-specific mods. Users with older hardware, such as the RTX 3090, can now target around 30 FPS in supported titles. This change expands the practical utility of next-generation rendering features to a wider range of existing hardware.

Andrew Sampson quantizes model for BGFX to enable frame generation on older hardware

The software update comes from Andrew Sampson, the developer behind the Borderless Gaming tool. The core innovation lies in how the application handles the DLSS 5 neural rendering model. Instead of relying on NVIDIA's official software stack, the app uses the BGFX compositor to capture and process frames. This method allows the game to run the model directly through BGFX rather than forwarding requests to NVIDIA's DLLs.

Sampson quantized the DLSS 5 model to make it compatible with unsupported GPUs. The effect is not bundled by default with the application. Users must install a single shader file from a GitHub repository. The installation process involves pasting the repository URL to download the necessary files. This approach bypasses the need to replace game DLLs or modify game files.

The tool provides simplified controls for strength, lighting, and structure. This offers a more accessible experience compared to complex injection tools that require hunting for game-specific mods. While the RTX 3090 achieves around 30 FPS, performance on other non-RTX 50 series GPUs remains unconfirmed. The update focuses on ease of use and broader hardware compatibility for frame generation.

This update allows Borderless Gaming users to apply DLSS 5 Frame Generation without official NVIDIA support. The BGFX-based implementation provides a viable alternative to traditional injection methods. Gamers with unsupported cards can now access these features through a straightforward GitHub installation.