This good thing. I gave it to others.

After AMD launched FSR Redstone at the last Computex 2025 event, this latest round is not just the upscaling of previous FSR portraits, but a step towards Neural Rendering technology that uses Machine Learning to enhance both graphic quality and serious gaming performance.

.

What surprised many people is that AMD is not only available on Radeon cards, but also supports NVIDIA GeForce and Intel Arc. It uses the ML2CODE system that converts trained ML models into Shader Code that can be run on almost any GPU.

.

ML2CODE is the core of Redstone because it converts the Neural Rendering Core into standard code. Whether it's DirectX HLSL or Vulkan GLSL, it can be processed immediately, eliminating the need to lock in one architecture.

.

In other words, if it is AMD's open source HIP Code, it can be compiled into CUDA, run on NVIDIA cards, or used with Intel Arc Shader. This technology has become a cross-platform solution that developers code once, but all screen card camps have the same access.

.

Another interesting point is that FSR Redstone does not require AI Core to be used. Chris Hall of AMD confirms that older cards without dedicated AI processors can still be used because the system switches ML to Shader Code and sends them to the GPU to run directly. Even with a slight increase in Overhead, it does not even eliminate the opportunity for older users, unlike FSR 4, which is restricted to RDNA 4.

.

This kind of openness changes the game on both the user and the developer side. New games can unlock a more beautiful and slippery graphic experience without being tied to any camp hardware. Players using previous cards like RDNA 3 or even the GeForce RTX 30 Series also have the opportunity to try something new. It's called FSR Redstone. It's not just an AMD feature, it's becoming a new standard in the gaming industry.

.

Source: wccftech

# IT News # Includes IT matters # Cough to know # IT