Eric Heitz retweeted
Genesis World 1.0 is out!
About 10 months ago I joined Genesis AI and have since been leading the rendering team, building Nyx, a renderer purpose-built for robotics.
Watch the video. Real or sim? When does it transition?
We are back again :) After three weeks of quiet building.
Introducing Genesis World 1.0, our latest simulation platform, the second release in our full-stack suite. Open-sourced.
Robotics is still bottlenecked by the 1× speed of the physical world. Every model, checkpoint, and data recipe eventually needs to be tested on physical hardware, slowly, expensively, and with limited coverage.
One hour in reality can become 100 days in simulation. That is how robotics model iteration moves from a wall-clock bottleneck to a compute problem.
To make this work, simulation has to be both fast and trustworthy.
Over the past year, we rebuilt the entire stack: a GPU-accelerated cross-platform compiler, penetration-free multi-physics contact solvers, unified rigid and deformable physics, and a photo-realistic renderer purpose-built for physical AI applications.
We built Nyx, a high-performance path-traced rendering engine for robotics application.
Genesis World 1.0 achieves near realtime performance with our latest development for penetration-free IPC solver, supporting various types of deformables beyond rigid bodies. It supports contact-rich, dexterous manipulation simulation across different embodiments: unitree, sharpa, wuji, genesis hand and various types of grippers.
Under the hood is Quadrants, our effort in pushing forward cross-platform GPU-accelerated computation. Quadrants started as a fork of Taichi, and we rebuilt most of the critical parts for optimizing simulation workloads, giving 10x faster launch time and up to 4.6x runtime performance compared to the initial Genesis release.
Together, they bring us to an unprecedentedly low sim-to-real gap, enabling zero-shot real-to-sim model evaluation and much faster iteration of GENE.
All available today.
Genesis World 1.0: github.com/Genesis-Embodied-…
Quadrants: github.com/Genesis-Embodied-…
Nyx: github.com/Genesis-Embodied-…
Eric Heitz retweeted
So... Textured Area Lights are ready to be reviewed - I hope they can make it into #Godot 4.7 🙈
I was able to integrate @eric_heitz's area light texture sampling into the forward renderer, lightmaps volumetric fog, and voxelGI.
#ScreenshotSaturday #GodotEngine #GameDev
Eric Heitz retweeted
A Fluorescent Material Model for Non-Spectral Editing & Rendering
Belcour Laurent @_Laurent , Fichet Alban, Barla Pascal
(Intel Corp. France, Inria Bordeaux)
arxiv.org/abs/2505.19672
Abstract:
Fluorescent materials are characterized by a spectral reradiation toward longer wavelengths. Recent work [Fichet et al. 2024] has shown that the rendering of fluorescence in a non-spectral engine is possible through the use of appropriate reduced reradiation matrices. But the approach has limited expressivity, as it requires the storage of one reduced matrix per fluorescent material, and only works with measured fluorescent assets.
In this work, we introduce an analytical approach to the editing and rendering of fluorescence in a non-spectral engine. It is based on a decomposition of the reduced reradiation matrix, and an analytically-integrable Gaussian-based model of the fluorescent component. The model reproduces the appearance of fluorescent materials accurately, especially with the addition of a UV basis. Most importantly, it grants variations of fluorescent material parameters in real-time, either for the editing of fluorescent materials, or for the dynamic spatial variation of fluorescence properties across object surfaces. A simplified one-Gaussian fluorescence model even allows for the artist-friendly creation of plausible fluorescent materials from scratch, requiring only a few reflectance colors as input.
Eric Heitz retweeted
Happy to share that our paper with @_Laurent, "Hardware Accelerated Neural Block Texture Compression with Cooperative Vectors," was accepted at HPG'25!
We dive deeper into the complete subject. I'll be presenting it later this month, see you there!
🔗 ggx-research.github.io/publi…
Eric Heitz retweeted
Microsoft has officially launched the Coop Vectors extension for DX12, enabling the use of matrix mul engines across vendors in a convinient way.
This opens the door for a range of new rendering algorithms, making them more practical for game dev.
devblogs.microsoft.com/direc…
Eric Heitz retweeted
I presented at MSFT’s AGS a demo that used coop vectors for NTC in smth representative of a modern RP with real texture contraints: size, quality, compression, filtering, inference overhead in ms.
The talk will soon be in the GDC Vault, in the meantime:
devblogs.microsoft.com/direc…
Eric Heitz retweeted
If you happen to be at #GDC2025, I'll be giving a talk about Neural Texture Compression and Cooperative Vectors at Microsoft's Advanced Graphics Summit, feel free to attend!
schedule.gdconf.com/session/…
Eric Heitz retweeted
“EON: A practical energy-preserving rough diffuse BRDF”
Jamie Portsmouth, Peter Kutz, Stephen Hill
arxiv.org/abs/2410.18026
(GLSL code in paper)
Abstract:
We introduce the "Energy-preserving Oren--Nayar" (EON) model for reflection from rough surfaces. Unlike the popular qualitative Oren--Nayar model (QON) and its variants, our model is energy-preserving via analytical energy compensation. We include self-contained GLSL source code for efficient evaluation of the new model and importance sampling based on a novel technique we term "Clipped Linearly Transformed Cosine" (CLTC) sampling.
#rendering
Eric Heitz retweeted
I can't believe we're done!
Time Ghost
A short film we on the Demo Team have been working on for a while
youtu.be/o1JIK5W3DRU
Like all our team’s demos, it runs in real-time at 30fps in Unity. Over the next days I’ll be posting rendering and tech art breakdowns in this 🧵
Eric Heitz retweeted
Looking for a fast method to render your large-scale game components?
Our lastest @HPG_Conf paper allows to render terrains, oceans and even entire planets in realtime on any modern GPU using compute shaders. Fun fact: we use a Concurrent Binary Tree as a GPU memory manager.
Eric Heitz retweeted
Can't wait for the presentation?
Here is our project webpage: ggx-research.github.io/publi…
Eric Heitz retweeted
I'm excited to share our #SIGGRAPH2024 paper on unifying stochastic representations for light transport:
cs.dartmouth.edu/~wjarosz/pu…
We show how to use stochastic implicit surfaces to describe microfacet surfaces and participating media (and a bunch of new things in between).
Eric Heitz retweeted
Time to share our #SIGGRAPH2024 paper "N-Dimensional Gaussians for Fitting of High Dimensional Functions"!
We show that N-Dimensional Gaussian mixtures can be optimized to handle high dimensional inputs like MLPs with a fraction of the training times.
sdiolatz.info/ndg-fitting/
Our I3D paper won the best paper presentation & the best paper runner-up awards. Congrats on the amazing presentation @thomasdeliot !
In our latest @I3DCONF paper, we transform any rasterizer into a differentiable one with minimal engineering and zero external dependencies.
Optimize game assets and volumes, train 3D Gaussian Splats, etc. directly within your existing engine. Compute shaders is all you need!
Eric Heitz retweeted
A minimalist unity sample for optimizing a triangle soup is now live on the project page :)
ggx-research.github.io/publi…
In our latest @I3DCONF paper, we transform any rasterizer into a differentiable one with minimal engineering and zero external dependencies.
Optimize game assets and volumes, train 3D Gaussian Splats, etc. directly within your existing engine. Compute shaders is all you need!
By implementing our method in @unity, we can achieve fully cross-platform differentiable rendering, running even on low-end mobile phones:
Eric Heitz retweeted
We wrote a tiny blogpost about a nice trick to speed up VNDF importance sampling an isotropic Smith-GGX distribution. You can find it here, hope you find it useful: auzaiffe.wordpress.com/2024/…

























