Linux × AMD: Patches verbessern Raytracing-FPS um bis zu 14 Prozent https://www.computerbase.de/news/betriebssysteme/linux-x-amd-patches-verbessern-raytracing-fps-um-bis-zu-14-prozent.93534/ #mesa #amd #raytracing #linux

It's funny how simple and fast a basic raytracer can be.

I remember in my first year of IT school (2007-2008) when we had to implement one on CPU, and we had to wait several minutes for one picture to be rendered.

Of course this one is still very basic, but hey it's in real time baby

Tiny distance to sphere in #glsl

(radius of 1, no positioning, and don't be inside the sphere, but who cares about all that)

New paper on my blog (available in English):

https://antoninjousse.com/BLOG/ENGLISH/ARTICLES/005-sunset-carla-island-EN.html

Nouvel article sur mon blog :
"The Sunset of Carla’s Island"

https://antoninjousse.com/BLOG/ARTICLES/005-sunset-carla-island.html

#research #blog #digital #numérique #carla #island #virtuel #nelson #3d #raytracing #beach

_

POV-Ray animation of Lorenz attractor

https://youtu.be/6y7HPqlL3J4?si=P8OvqQMXjTIuCbNT

#POVray #raytracing
_

I used to pick the time as the 3rd dimension to debug the 3D gradient noise, but unless you're into cloudy atmospheres, it's much better to display it on a 3D sphere.

Note: partial derivative magnitude on the right side if you wonder

I've been playing with #povray #raytracing today, for the first time in a while. Here are the beginnings of a scene that will serve as a manual cover and (suitably converted) a title screen for some future game.

Update on my #Raytracing in One Weekend implementation in #Zig: I gotta lotta spheres. Also implemented multithreading.

Examining #RDNA4's out-of-order memory accesses in detail, and investigating with testing
#RDNA 4's memory subsystem enhancements are exciting and improve performance across a variety of workloads. #AMD specifically calls out benefits in #raytracing. But RDNA 4's scheme for handling memory dependencies isn't fundamentally different from that of #GCN many years ago. RDNA 4’s makes the most significant change to AMD’s #GPU memory subsystem since RDNA launched in 2019.
https://chipsandcheese.com/p/rdna-4s-out-of-order-memory-accesses

#Raytracing #BVH traversal struggles on GPUs due to irregular control flow.

Our Locally-Subdivided Neural Intersection Function (LSNIF) is an ML-based alternative for optimized #rendering.

Join us at @i3d for our talk!

Learn more: https://gpuopen.com/learn/using_neural_networks_for_geometric_representation/?utm_source=mastodon&utm_medium=social&utm_campaign=lsnif

But what if it didn't have ray tracing?

Nintendo Confirms Switch 2 Uses DLSS and Ray Tracing, but Is Being Super Vague About the Details - IGN

https://www.ign.com/articles/nintendo-confirms-switch-2-uses-dlss-and-ray-tracing-but-is-being-super-vague-about-the-details

#raytracing folks, what do you think of next event estimation vs simple MIS (generate either brdf sample or light sample, use total probability in MC estimator)?

My former student insists on NEE, but it has so many special cases that I doubt it's a good idea pedagogically.

Just realised that in my software GPU wavefront pathtracing I could implement efficient virtual memory by storing any 'page miss' paths and then sort by page, load/unload pages and process the relevant paths.

Does anyone know if this has been done before?

#FluidX3D #CFD v3.2 is out! I've implemented the much requested #GPU summation for object force/torque; it's ~20x faster than #CPU #multithreading.
Horizontal sum in #OpenCL was a nice exercise - first local memory reduction and then hardware-supported atomic floating-point add in VRAM, in a single-stage kernel. Hammering atomics isn't too bad as each of the ~10-340 workgroups dispatched at a time does only a single atomic add.
Also improved volumetric #raytracing!
https://github.com/ProjectPhysX/FluidX3D/releases/tag/v3.2