Today we're going to talk about this presentation by Sergei Kulikov from Machine Games at SIGGRAPH 2025, where he details the hair rendering system implementation for the game Indiana Jones and the Great Circle, keeping the 60 FPS target, using asynchronous compute and LOD management to limit segment count to maximum 2 million.

Read full presentaion:

https://advances.realtimerendering.com/s2025/content/Strand%20Hair%20in%20IJGC%20-%20Final%20Slides%20(Post-Conference).pdf

Original Copyright Free Music written, performed and produced by TrackTribe:

https://www.youtube.com/watch?v=HK9z2cR8nlM&list=PLFEJ2E5exsjrjRzhJdOYOmTlcYfK3Hz7_

On this occation we're going to talk about this paper by Eugene D'Leon and Benedikt Bitterli from NVIDIA as well as Dario Seyb and Wojciech Jarosz from Dartmouth College, presenting a unified theory of light transport using Stochastic Implicit Surfaces, specifically focusing on Gaussian Process Implicit Surfaces (GPIS); avoiding the usual separation of scenes en either deterministic surfaces or participating media, which can create some practical problems.

Read full paper here:

https://cs.dartmouth.edu/~wjarosz/publications/seyb24from-small.pdf

Background using royalty free music, including:

"⁠George Street Shuffle (ISRC USUAN1300035)⁠" by Kevin MacLeod is licensed under ⁠CC BY 3.0⁠.

In this session, we will analyze a paper by Jaime Postmouth, Peter Kuts, and Stephen Hill, in which they present a new bidirectional reflectance distribution function (BRDF) based on Oren–Nayar, but energy-preserving. They also introduce a sophisticated importance sampling scheme called CLTC, which drastically reduces variance in ray tracing.

Read full paper here:

https://arxiv.org/pdf/2410.18026

Legal:

La música de fondo es libre de regalías, incluyendo:

"⁠⁠George Street Shuffle (ISRC USUAN1300035)⁠⁠" by Kevin MacLeod is licensed under ⁠⁠CC BY 3.0⁠⁠.

Today we’ll be talking about the presentation given by Peter Shirley from Activision at High Performance Graphics 2025. In it, he discussed the use of exponential Spherical Harmonics to replace the Split-Sum method, by re-encoding the environment map with a view-dependent factorization. This approach removes the need for a multi-level cube map, saving disk space and making the process noticeably faster — though, with some limitations

Watch full presentation:

https://www.youtube.com/watch?v=jN7FX5COASM

On this occasion, we will discuss about the presentation by Eric Wohllaib, Graphics Programmer at Sucker Punch Productions, gave at te Game Developers Conference, about how the procedural grass system implemented for the game Ghost of Tsushima was developed using Compute Shaders and Bézier Curves—covering everything from instance data generation and level-of-detail management, to the unified wind system and its interaction with the player.

watch full presentation here:

https://www.youtube.com/watch?v=Ibe1JBF5i5Y

Join us in this talk where we will discuss the presentation given by Kentaro Suzuki, Kenichiro Yasutomi, and Hajime Uchimura from Polyphony Digital during SIGGRAPH 2025, where they delve into the meticulous approach to rendering car materials used in Gran Turismo 7. Since they often go beyond the Rec. 709 standard, they require wide color gamut rendering and adaptive tone mapping for HDR displays.

Read full presentation:

https://blog.selfshadow.com/publications/s2025-shading-course/pdi/s2025_pbs_pdi_slides.pdf

Today we will discuss the talk given by Angelo Pesce from Roblox at High-Performance Graphics 2025, where he explores the landscape of real-time rendering, particularly within the video game industry. He evaluates how the approach to video game development has changed to date and analyzes the opportunities and challenges of emerging technologies, such as Generative AI, Cloud Computing, and others.

Watch full presentation:

https://www.youtube.com/watch?v=7kYLdrfyk84

This time we're going to talk about this article by Jarkko Lempiainen, from HypeHype Games, where he details the development of a new stochastic lighting algorithm for the HypeHype user-generated content platform, designed to efficiently implement dynamic local lights and shadows on a wide range of mobile devices; which needed to be scalable and intuitive for non-technical users like kids, avoiding the performance limitations of traditional rendering techniques.

Read full article:

https://www.linkedin.com/pulse/local-lights-hypehype-jarkko-lempiainen-bjzpf/

Today we’ll talk about the presentation given by Yana Mateeva and Marco Bouterse, both graphics programmers at Nixxes Software, during the 2024 Graphics Programming Conference. In it, they discuss the challenges of bringing a game as heavily optimized for PlayStation as Ghost of Tsushima to PC — mainly dealing with unified memory management, resource handling, and the adaptation of shaders that relied on Sony platform–specific features.

Watch the full presentation:
https://www.youtube.com/watch?v=tostM4-rK6o&list=TLGGXWlK_jZBJ1UxOTA4MjAyNQ&t=45s

Today we're going to chat about this presentation given by Erick Jansson, Senior Graphics Programmer of Remedy Entertainment during the Digital Dragons conference from 2024, where he talks about the GPU-Driven Renedering pipeline they implemented for Alan Wake 2. Focusing mainly on their use of Mesh Shaders and diving into how they implemented their the frustum culling and occlusion culling to optimize the scene from 200 million triangles to just 10 million, among other things.

Watch full presentation here:

https://www.youtube.com/watch?v=EtX7WnFhxtQ