AMD’s Radeon Rays 4.0 Announced as Closed Source, Then Mostly Opened Again After Backlash

AMD’s Radeon Rays 4.0 Announced as Closed Source, Then Mostly Opened Again After Backlash

4 years ago
Anonymous $-9GJQVHNr8

https://wccftech.com/amds-radeon-rays-4-0-announced-as-closed-source-then-mostly-opened-again-after-backlash/

Mere days after the relaunch of the GPUOpen program featuring new tools and an expanded FidelityFX suite, AMD also launched the 4.0 version of Radeon Rays.

The ray intersection acceleration library (formerly known as FireRays) is part of the AMD ProRender software suite. However, it could previously only run on the CPU, which was quite the limitation. Now, with the first RDNA2 AMD GPUs already confirmed to introduce hardware support for ray tracing, Radeon Rays 4.0 finally introduces BVH optimization specifically for GPU access alongside requiring one of the major low-level APIs: Microsoft's DirectX 12, Khronos' Vulkan, and Apple's Metal. It also supports Heterogeneous-Compute Interface for Portability (HIP), which is AMD's C++ parallel computing platform (the equivalent of NVIDIA's CUDA).

AMD’s Radeon Rays 4.0 Announced as Closed Source, Then Mostly Opened Again After Backlash

May 16, 2020, 6:18pm UTC
https://wccftech.com/amds-radeon-rays-4-0-announced-as-closed-source-then-mostly-opened-again-after-backlash/ > Mere days after the relaunch of the GPUOpen program featuring new tools and an expanded FidelityFX suite, AMD also launched the 4.0 version of Radeon Rays. > The ray intersection acceleration library (formerly known as FireRays) is part of the AMD ProRender software suite. However, it could previously only run on the CPU, which was quite the limitation. Now, with the first RDNA2 AMD GPUs already confirmed to introduce hardware support for ray tracing, Radeon Rays 4.0 finally introduces BVH optimization specifically for GPU access alongside requiring one of the major low-level APIs: Microsoft's DirectX 12, Khronos' Vulkan, and Apple's Metal. It also supports Heterogeneous-Compute Interface for Portability (HIP), which is AMD's C++ parallel computing platform (the equivalent of NVIDIA's CUDA).