The upcoming version of Unreal Engine, 4.22, will include several new features.
The most interesting addition for our audience is probably “Early Access” support for DirectX 12 Raytracing (DXR) on DirectX 12. This includes the low-level framework to cast and evaluate rays in shaders (although they don’t clarify whether that means written shaders, nodes for graph-based shaders, or both) as well as higher-level features that use DXR, such as area lights, soft shadows, and reflections. They have also added a denoiser for shadows, reflections, and ambient occlusion, which will improve image quality with lower sample counts.
If you remember NVIDIA’s RTX announcement, many of their first-party demos were built using Unreal Engine 4. This includes the Star Wars demo with the two Stormtroopers putting their feet in their mouths on an elevator with their boss. It makes sense that Epic would be relatively far along in RTX support, especially just before GDC.
A few other additions include Visual Studio 2019 support (although Visual Studio 2017 is still the default). The new Unreal Audio Engine is now enabled by default for new projects, which was a complete re-write of the original system that started a few years ago. The old audio system was a bit of a mess, and, worse, varied from platform to platform.
Unreal Engine 4.22 also (experimentally) opts-in to the much longer file and paths names that were introduced with the Windows 10 Anniversary Update. The previous limit was 260 characters for a full path, which was defined as MAX_PATH in Win32. I’m not sure what the new limit is, but I think it’s 32,767 characters after expansion. I could be wrong, though.
If you have the Epic Launcher installed, whether it’s for Unreal Engine, Fortnite, something from the Epic Store, Unreal Tournament 4, or whatever, then you can check out Unreal Engine 4.22 for free. (Royalties apply under certain circumstances… but, at that point, you are making money off of it.)
One would think that they
One would think that they have added Ray Tracing support for any and all render passes that could benifit from DXR Ray Tracing in games! The games developers would be able to pick and choose which Ray Traced render passes would best benefit that games Art and Imagery as well as textures/other assets within the limitations of MS’s DXR and the GPU’s hardrware based RT/AI ray tracing/Tensor Core trained AI denoiser or for non RTX/Turing GPUs a software/Shader(Shader core’s Math/ALU units) code path for both Ray generation and runing the AI Denoiser.
I’m sure that MS is going to want to have as much GPU hardware able to make use of its DXR feature set as possible. So for GPUs Any Nvidia/All AMD GPUs that currently do not have have the RT cores/AI cores IP that Nvidia’s RTX branded GPU hardware offers, that will be some DXR software code path for DXR ray tracing in a more limited fashon for games and older GPU hardware.
And even Nvidia’s RTX Turing SKUs that offer 10 GigaRays/s performance that’s still only 10,000,000 rays per millesecond and, for example, that 10 million multiplied by 16.67ms(60 FPS Frame Time) that comes to 166,700,000 rays available for one Frame at 60 FPS. And that’s still going to be fruther time limited because in that time slot of 16.67ms all the other render pass mixing and denoising pipelines need to have their work performed also! And all that mixing of pipline results is not a fully parallelizable process thare are some dependencies there! So that’s even less time for Ray Casting available within the alloted 16.67ms time slot per frame-time limit for all that other passes/mixing/denoising to be completed.
So any denoising AI is going to help with the grainy images produced because of the limited amount of rays available per milleseconds of the various frame times between 30 FPS(33.33ms) and 60 FPS(16.67ms) and higher frame rates than 60 FPS that have frame times below 16.67ms.
I’d like to see how many rays/s could be generated on Pascal/eariler, or an AMD, GPU’s available shader cores using a Software/Shader route for ray calculations, ditto for AI denoising, in DXR compared to RTX Turings specilized Ray Tracing Cores and Tensor Cores hosted AI denoiser.
There really needs to be in the order of 1 more magnitude in the numbers of rays/s generated to 100 GigaRays/s on some Turing replacment to cover all the various passes that need rays for their stages to be completed in say 16.67ms/under frame times for high frame rate gaming.
Nvidia employs a lot of very creative short cuts in its RTX Hybrid “Real Time” Ray Tracing mixed with rasterization pipeline to get things done using the limited resources at hand and so did Imagination Technologies with their PowerVR Wizard Hybrid “Real Time” Ray Tracing mixed with rasterization Hardware based Ray Tracing IP.
Also how will the Vega 20 Based Radeon VII’s extended Vega-2 micro-arch based AI ISA extentions factor into any AI based denoising of some more limited numbers of available Rays that are calculated on Radeon VII’s shader cores. Radeon VII’s nCUs will be needed for the running the trained denoising AI also whereas Nvidia has specilized Tensor Core for that in addition to Turing’s RT cores.
Nvidia is fruther ahead of AMD this generation than previously.