Today, AMD launched its long awaited FSR Redstone update, which includes FSR Upscaling (formerly known as FSR 4), FSR Frame Gen, FSR Ray Regeneration, and FSR Radiance Caching.
The first two are improved versions of previous technologies. The third one, Ray Regeneration, was actually already introduced by Call of Duty: Black Ops 7. Developed in partnership with Microsoft, Ray Regeneration is a machine learning-powered denoiser that focuses on restoring details to ray traced effects, which can be troublesome in upscaled scenes. It is, essentially, AMD's version of NVIDIA DLSS Ray Reconstruction.
However, FSR Radiance Caching is not yet available in any game. During today's presentation, AMD showcased a tech demo of the feature in Warhammer 40K Darktide, Fatshark's first-person online action game that just got the new Hive Scum class DLC. Viewers assumed this meant Darktide would also be the first game to implement FSR Radiance Caching, but in a press release, the Swedish developer clarified that the tech is 'experimental' and therefore not currently planned for release in Darktide's live build, as it requires 'further development, optimization, and quality verification before it could be integrated into a live service game.'
That said, Fatshark was eager to jump at the opportunity to develop the tech demo when AMD approached them about it this past Summer. Their Chief Technology Officer, Mikael Hansson, said in a statement:
Technology is meant to be built and shared together. Working with innovative powerhouses like AMD keeps us sharp, encourages diversification in game technology, and strengthens the entire industry. It’s fun, it’s challenging, and it’s a long-term investment, not just in Fatshark’s future, but in the future of gaming.
In the newly released FSR Redstone SDK, AMD describes FSR Radiance Caching as a 'state-of-the-art illumination cache designed to work in tandem with Monte Carlo path tracing to boost rendering performance. At its core is an online machine learning model that continuously trains on complex, multi-bounce global illumination (GI) on-device and in real-time. The result is richer, more immersive lighting environments that respond dynamically without the need for pre-computation or baking.'
It sounds fairly similar to NVIDIA's RTX Neural Radiance Cache, another technique that aims to make path tracing faster with the help of machine learning by inferring an infinite amount of bounces after the initial one or two bounces from path traced rays. RTX Neural Radiance Cache was released through the RTX Global Illumination SDK at CES 2025, though it is only available in Portal with RTX so far.
About the author: With over two decades of experience in gaming journalism, Alessio Palumbo has led the gaming vertical at Wccftech since August 2015. He started working at a young age for Italian websites like Everyeye.it, Gamestar.it, Nextgame.it, and Multiplayer.it before kickstarting the indie English-language publication Worlds Factory as its founder and Editor in Chief. In the last decade, he has coordinated the overall output of Wccftech's gaming section, managed PR relations, assigned reviews, produced daily news coverage, edited gaming content as needed, and delivered game reviews. Arguably, his trademark content is the long series of exclusive developer interviews that have been cited by Wikipedia and by the biggest news media and gaming publications. His passion for technology also makes him knowledgeable when it comes to gaming hardware and tech. His favorite genres include RPGs, MMORPGs, and action/adventure games.
Follow Wccftech on Google to get more of our news coverage in your feeds.
