Borderless Gaming Brings DLSS 5-Style Neural Rendering to Unsupported GPUs

Borderless Gaming has received a notable update that brings an unofficial implementation of NVIDIA's DLSS 5 Neural Rendering technology to a much wider range of graphics cards.

Unlike other experimental DLSS 5 solutions that require replacing game files or injecting NVIDIA DLLs, this implementation works through the application's existing BGFX graphics pipeline. The developer has also reworked and quantized the neural-rendering model to reduce its hardware requirements.

DLSS 5 Without Modifying Game Files

One of the biggest differences between this approach and previous DLSS 5 experiments is its simplicity.

The Borderless Gaming implementation does not require users to replace DLL files, modify game installations or install a separate game-specific modification. Instead, the effect can be installed separately through the application's effects system and applied to compatible games.

This makes the technology considerably easier to experiment with than some earlier unofficial implementations.

It Doesn't Use NVIDIA's DLSS Software

Despite being referred to as a DLSS 5 implementation, the Borderless Gaming version does not directly call NVIDIA's DLSS software or DLLs.

According to developer Andrew Sampson, the neural-rendering model was extensively reworked, including quantization and the removal of processing bottlenecks. The modified model runs directly through BGFX, rather than sending requests through NVIDIA's software stack.

Because of this, it is more accurate to describe the project as a DLSS 5-derived or DLSS 5-style implementation, rather than native NVIDIA DLSS 5 support.

Wider GPU Compatibility

NVIDIA's official DLSS 5 Neural Rendering support is currently focused on GeForce RTX 50-series GPUs.

The Borderless Gaming project takes a different approach by adapting the model so it can operate on a broader selection of hardware.

The developer claims that even an RTX 3090 can target around 30 frames per second with the implementation, although actual performance will vary depending on the game, resolution and selected settings.

The broader compatibility could make neural rendering accessible to users who do not own NVIDIA's latest graphics cards.

Adjustable Neural Rendering Controls

The new effect includes several settings that allow users to adjust how the neural-rendering process works.

Available controls include:

• Rendering strength
• Lighting
• Structure
• Style
• Processing resolution

These options give users some control over the visual effect and its performance requirements. Lower processing resolutions can potentially reduce the workload, while higher settings can increase the processing demand.

How the Borderless Gaming Method Works

The implementation uses BGFX, which allows Borderless Gaming to capture the game window at the compositor level and process it through a GPU-based effect pipeline.

This means the effect can be applied without requiring developers to integrate DLSS 5 directly into a game's rendering engine.

The approach is therefore different from NVIDIA's native implementation, which is integrated into the game's rendering pipeline and can make use of information such as rendered frames and motion vectors.

The DLSS 5 Effect Is Optional

The developer has emphasized that the DLSS 5-related shader is not automatically installed with the Borderless Gaming update.

Users must manually install the effect if they want to experiment with it. The application provides a system for installing effects and presets from GitHub by entering a repository URL.

This means simply updating Borderless Gaming does not automatically activate the neural-rendering effect or consume GPU resources in the background.

Performance Could Improve Further

The current implementation is still experimental.

Borderless Gaming's developer says the build uses experimental shader technology and expects performance to improve as graphics drivers gain better support for Shader Model 6.10.

Performance will also depend heavily on the GPU, game, resolution and processing settings used by the player.

As a result, users should not expect the same performance or image quality as NVIDIA's official DLSS 5 implementation on supported hardware.

A Growing DLSS 5 Modding Scene

This latest development is part of a growing community effort to bring DLSS 5 Neural Rendering beyond its official hardware limitations.

Recent experiments have demonstrated DLSS 5-related neural rendering on older NVIDIA GPUs, AMD Radeon graphics cards and even Intel hardware. Some projects have also explored open-source Vulkan implementations.

These projects are technically interesting because developers are finding alternative ways to reproduce or adapt neural-rendering workloads across different GPU architectures.

Official DLSS 5 vs Borderless Gaming Implementation

What This Means for PC Gamers

The Borderless Gaming implementation provides another way for PC gamers to experiment with neural rendering without purchasing a newer GPU or modifying individual game installations.

However, users should remember that this is not official NVIDIA DLSS 5 support. Image quality, compatibility and performance can vary considerably between games and hardware.

The project is also still evolving, so future updates could improve performance and expand compatibility.

Final Thoughts

Borderless Gaming's latest update demonstrates how quickly the PC gaming community is experimenting with NVIDIA's new neural-rendering technology.

By quantizing and restructuring the model and running it through BGFX, the developer has created an implementation that can operate on hardware beyond NVIDIA's officially supported RTX 50 series.

For now, it remains an experimental project rather than a replacement for native DLSS 5. Nevertheless, its wider hardware compatibility and lack of game-file modification make it an interesting development for PC gamers interested in experimenting with neural rendering.