Game development presentations - a useful reference

[jlippo]

Machine Learning Summit: Real-Time Neural Texture Upsampling in 'God of War Ragnarok'

This talk gives a detailed breakdown of the run-time neural texture upsampling system in God of War Ragnarkincluding how the networks are designed to directly output BC compressed image blocks, how the process is optimized to take full advantage...

www.gdcvault.com

EDIT:

Nice to see someone try raytraced cubemaps, been wondering about the idea for a while.

Upsampling was interesting as well.
Reminded about the bicubic upsampling of virtual textures in Rage.
Should work nicely in such case.

 

[chris1515]

Picking a Good Spot: Naughty Dog's Post System

In this talk, Game Programmer Ming-Lun

gdcvault.com

Naughty Dog engine GDC 2023 talk

Maximum supported posts went from 8,000 (SP) to hundreds of thousands (plus a lot more NPCs) due to having "much larger levels".

 

[chris1515]

 

[chris1515]

 

[Remij]

That was a fun one!

 

[chris1515]

 

[chris1515]

 

[chris1515]

 

[SlmDnk]

Random-Access Neural Compression of Material Textures - NVIDIA

The continuous advancement of photorealism in rendering is accompanied by a growth in texture data and, consequently, increasing storage and memory demands. To address this issue, we propose a novel neural compression technique specifically designed for material textures. We unlock two more levels of detail, i.e., 16X more texels, using low bitrate compression, with image quality that is better than advanced image compression techniques, such as AVIF and JPEG XL. At the same time, our method allows on-demand, real-time decompression with random access similar to block texture compression on GPUs, enabling compression on disk and memory. The key idea behind our approach is compressing multiple material textures and their mipmap chains together, and using a small neural network, that is optimized for each material, to decompress them. Finally, we use a custom training implementation to achieve practical compression speeds, whose performance surpasses that of general frameworks, like PyTorch, by an order of magnitude.

Video: https://research.nvidia.com/labs/rtr/neural_texture_compression/assets/ntc_video.mp4
Image viewer: https://research.nvidia.com/labs/rtr/neural_texture_compression/assets/Results/results.html


Real-Time Neural Appearance Models - NVIDIA

We present a complete system for real-time rendering of scenes with complex appearance previously reserved for offline use. This is achieved with a combination of algorithmic and system level innovations.

Video: https://research.nvidia.com/labs/rt...ets/nvidia_neural_materials_video-2023-05.mp4
Image viewer: https://research.nvidia.com/labs/rtr/neural_appearance_models/imageviewer/imageviewer.html

 

[chris1515]

 

[Scott_Arm]

Introducing AMD Capsaicin Framework - our ARR Research Framework which includes our GI-1.0 sample

AMD Capsaicin Framework is a research framework for real-time graphics which we use to develop and test new and upcoming rendering algorithms.

gpuopen.com

GitHub - GPUOpen-LibrariesAndSDKs/Capsaicin: AMD ARR team rendering framework

AMD ARR team rendering framework. Contribute to GPUOpen-LibrariesAndSDKs/Capsaicin development by creating an account on GitHub.

github.com

AMD's new rendering framework released, including GI-1.0 as described in the youtube video in the post above.

 

[JoeJ]

Raytracing using AMD HIP: https://gpuopen.com/hiprt/

Can build offline BVH:

HIP RT can then be used to perform the intersection on the GPU. As of today, this is a unique feature of HIP RT that no other ray tracing APIs support. For example, HIP RT can be used to build a high quality BVH for static geometry in a scene and then load it at runtime.

So i guess intersection instructions are exposed as well, but i could not find the actual implementation code, just samples.
With HIP now coming to Windows, maybe that's an option for me to make traceable LOD... : )

 

[DegustatoR]

https://twitter.com/x/status/1656649916247166976

 

[chris1515]

 

[chris1515]

Learn Game Engine Programming

engine-programming.github.io

A great blog about writing a game engine

 

[chris1515]

 

[chris1515]

 

[DegustatoR]

Interactive Hair Simulation on the GPU Using ADMM | Research

We devise a local–global solver dedicated to the simulation of Discrete Elastic Rods (DER) with Coulomb friction that can fully leverage the massively parallel compute capabilities of moderns GPUs. We verify that our simulator can reproduce analytical results on recently published cantilever...

research.nvidia.com

 

[Karamazov]

 

[Remij]

Looking forward to seeing Sebbi's talk "Modern Mobile Rendering @ HypeHype" at REAC (Rendering Engine Architecture Conference)

I don't know when the presentation will be posted, but it is being recorded.. so I'm definitely looking forward to it.

Rendering Engine Architecture Conference

A conference about engine architecture for real-time renderers

enginearchitecture.realtimerendering.com

His twitter feed is a treasure trove of insight into modern rendering. If you're new to the forum, he used to post here, and has done lots of amazing work in his career.

https://twitter.com/SebAaltonen