Intel Arc FSR enabled with OptiScaler
The quest for seamless upscaling in graphics is a perpetual game of technological catch-up. Recently, the modding community has delivered a fascinating breakthrough, successfully running AMD’s advanced FSR 4.1.2 upscaler directly on Intel Arc graphics cards, unlocking performance previously thought exclusive to other architectures.
This achievement isn’t just a technical feat; it’s a triumph of community innovation. By leveraging dedicated XMX acceleration hardware and modifying INT8 DLL files, developers have bypassed typical fallback paths to route the neural upscaling models directly through Intel’s Xe Matrix Extensions. This ingenious workaround allows the Arc architecture to benefit from AMD’s cutting-edge upscaling technology, delivering performance that rivals Intel’s native XeSS Super Resolution.
While AMD has expanded official FSR 4.1 support for Radeon RX 6000 cards and Ryzen AI processors, running the feature on Intel silicon has remained an unsanctioned community effort. This independent modification demonstrates the immense creative drive within the hardware enthusiast sphere.
The results speak for themselves. Early tests on cards like the Arc B580 and Arc A770 show remarkable efficiency. In demanding scenarios, the FSR 4.1.2 Lite model demonstrated incredible speed. For instance, in the game Clair Obscur: Expedition 33 at 1080p, the upscaling process took just 0.59ms, narrowly beating the 0.58ms recorded by XeSS Super Resolution, and significantly outperforming the full FSR 4.1.2 model’s 0.85ms.
The performance leap continued in graphically intensive titles. When tested with The Witcher 3: Wild Hunt Remastered at 1080p Quality on an Arc A770, the Lite FSR 4.1.2 model managed 101.5FPS, closely matching XeSS’s 101.1FPS, while the full upscaler delivered 98.6FPS.
Beyond raw frame rates, testers noted subjective improvements in visual quality. Enthusiasts reported that the streamlined Lite model not only offered similar performance but also successfully preserved much of the full version’s reconstruction clarity, suggesting a sophisticated balance between speed and visual fidelity.
Although this setup remains experimental and requires manual file replacements, targeting DirectX 12 games, the ingenuity behind it is undeniable. It highlights how dedicated hardware and community collaboration can push the boundaries of what different GPU architectures can achieve, setting a new standard for cross-platform performance.
Credit
