Experimental DLSS 4.5 Support Arrives for AMD Radeon RX 9000 Series
A breakthrough in graphics technology has emerged as the developer known as “countervolts” released a significant update to the d4r project, enabling experimental support for NVIDIA DLSS 4 and 4.5 technologies on AMD Radeon RX 9000 series graphics cards. By leveraging a Linux and Proton-based software stack, this project successfully integrates NVIDIA’s proprietary libraries with AMD hardware. This initiative specifically targets the native FP8 calculation capabilities of the RDNA 4 architecture, marking a pivotal technical step for users seeking to utilize NVIDIA’s advanced image scaling features on non-NVIDIA hardware. While currently in a proof-of-concept phase, the project offers a glimpse into a future where vendor-locked technologies might achieve broader cross-platform compatibility.
- The d4r project enables NVIDIA DLSS 4 and 4.5 functionality on AMD Radeon RX 9000 series cards featuring RDNA 4 architecture.
- RDNA 4 architecture provides native support for FP8 calculations, which effectively minimizes performance overhead during the DLSS upscaling process.
- The software utilizes ZLUDA to translate CUDA workloads for execution within Linux and Proton environments.
- The current iteration remains in a preliminary development phase and does not yet offer the stability required for standard daily gaming usage.
DLSS 4.5 Technology Achieves Compatibility with RDNA 4 Architecture
NVIDIA’s DLSS suite has traditionally remained exclusive to GeForce RTX hardware, yet the d4r project continues to challenge these established industry boundaries. The latest updates are specifically optimized for RDNA 4-based GPUs, such as the RX 9060 and 9070 models, allowing for the generation of custom kernels tailored to this hardware.
One of the most notable aspects of this integration is the ability of the RDNA 4 architecture to process the FP8 calculations required by modern DLSS models natively. This eliminates the need for expensive conversion processes, which significantly lightens the processing load and contributes to smoother frame rates.
However, users should exercise caution, as the project is still in its infancy. Although the documentation lists high-end cards like the RX 9070 XT as target hardware, developers maintain that real-world testing remains extremely limited. The experimental nature of this software means that while the technical foundation is present, the final user experience is subject to frequent bugs and graphical artifacts.
Development Process Proceeds Under Controlled Conditions
The core mechanism behind the d4r project involves rerouting the standard NVIDIA nvngx_dlss.dll file to operate on AMD GPUs via the ZLUDA translation layer. By introducing specialized kernels for RDNA 3 and RDNA 4 architectures, the developers have successfully mitigated much of the performance loss typically associated with such software-based translation methods. Tests conducted on the RX 7700 XT have already demonstrated higher frame rates compared to basic, unoptimized translation attempts.
Despite these promising technical milestones, this project is not intended to replace AMD’s proprietary FidelityFX Super Resolution (FSR) technology. The current lack of verified performance data and the experimental framework categorize this tool as a resource for technical enthusiasts rather than mainstream consumers. The community continues to monitor whether future updates will translate these experimental findings into a more robust and stable implementation for wider use. The ongoing evolution of this software raises interesting questions about the future of hardware-agnostic upscaling solutions in the gaming industry.
What are your thoughts on running NVIDIA’s DLSS technology on AMD hardware through experimental projects like d4r? Do you believe these developments should pressure AMD to accelerate the evolution of their own FSR technology, or is a more open ecosystem ultimately better for the gaming community? We invite you to share your perspectives and technical insights in the comments section below.
Your comment has been submitted,
it will be published after approval.