CUDA Ray Tracing 2x Faster Than RTX: My CUDA Ray Tracing Journey

https://karimsayedre.github.io/RTIOW.html

71•ibobev•7mo ago

Comments

sheepscreek•7mo ago

I’m guessing it’s because they’re using all the computing power the GPU has to offer in CUDA mode, as opposed to sharing the GPU with other functions (when in RTX).

colechristensen•7mo ago

Yup this is an "assume spherical cow" situation where it's not dishonest, but you can't draw any real world conclusions from the experiment unless you happen to be working in a very restricted space.

ChocolateGod•7mo ago

Wouldn't you need to in a real world scenario make the CUDA cores aware of the game geometry adding more work on the CPU?

touisteur•7mo ago

Ideally you don't make the cuda cores aware but rather the ray-tracing circuitry. RT cores are designed to perform ray-triangle intersections in a BVH. You get the teraflops and memory bandwidth (or more of it) if you fit the RT-core computing model.

And in most cases it's ok to spend time on one CPU function (creating and loading the BVH) against the hundred thousands of frames you'll be drawing on GPU.

colechristensen•7mo ago

A whole lot of stuff is going on during gaming and graphics rendering with trick upon trick to squeeze out every last bit of performance. Unless you're an expert in a graphics rendering stack or a game engine it's hard to have these conversations in a meaningful way.

atq2119•7mo ago

More likely it's because the scene they're using is completely unrepresentative of what people are interested in: almost no triangles, primarily procedural nodes (for spheres), and in general a fairly simple scene.

E-Reverance•7mo ago

See comments as to why this is misleading https://www.reddit.com/r/GraphicsProgramming/comments/1ljmf0...

pixelpoet•7mo ago

> FMA performance here is a non-issue, I'm not just flexing—I'm showing off my CUDA prowess. But hey, got to demonstrate I know my hardware!

This article is pretty embarrassing, and as others have noted, very misleading due to the RTX units hardly being used.

kachapopopow•7mo ago

wow, bypassing a rendering backend makes things go faster, what a surprise!

This only runs on nvidia, vulkan is designed to be cross-compatible with not only gpus, but operating systems as well. Vulkan is pretty direct compared to something like dx11 thought so I guess it is interesting to see performance improvement non the less.

esperent•7mo ago

I think the title of this should be changed, at the moment it's click bait. It should be something like:

CUDA Ray Tracing 2x Faster Than RTX when Rendering Spheres

As far as I can see, this renderer can't do anything else except spheres (and maybe planes).

It's no bad achievement to beat a general purpose production renderer at one specific thing, but a renderer that can only do spheres is just a hyper-optimized toy, and here it's being presented as far more than that.

kookamamie•7mo ago

> __restrict__ Pointers

Ahh, my favorite nitpick from C++ not having sane default aliasing rules spills to the CUDA-land.

pjmlp•7mo ago

Is hard to have them, when one of the original goals was being mostly copy paste compatible with C89.

kookamamie•7mo ago

Yes, though C has restrict in the language now, but C++ does not.

pjmlp•7mo ago

Because no one has ever bothered to create a WG21 paper proposal to include it.

We Mourn Our Craft

Speed up responses with fast mode

Hoot: Scheme on WebAssembly

U.S. Jobs Disappear at Fastest January Pace Since Great Recession

Stories from 25 Years of Software Development

OpenCiv3: Open-source, cross-platform reimagining of Civilization III

Al Lowe on model trains, funny deaths and working with Disney

The AI boom is causing shortages everywhere else

The Waymo World Model

Reinforcement Learning from Human Feedback

Start all of your commands with a comma (2009)

Vocal Guide – belt sing without killing yourself

France's homegrown open source online office suite

Coding agents have replaced every framework I used

Selection Rather Than Prediction

A Fresh Look at IBM 3270 Information Display System

72M Points of Interest

Show HN: I saw this cool navigation reveal, so I made a simple HTML+CSS version

I Write Games in C (yes, C)

Unseen Footage of Atari Battlezone Arcade Cabinet Production

SectorC: A C Compiler in 512 bytes

Where did all the starships go?

Software factories and the agentic moment

Show HN: Look Ma, No Linux: Shell, App Installer, Vi, Cc on ESP32-S3 / BreezyBox

Learning from context is harder than we thought

Monty: A minimal, secure Python interpreter written in Rust for use by AI

Making geo joins faster with H3 indexes

Show HN: Kappal – CLI to Run Docker Compose YML on Kubernetes for Local Dev

Hackers (1995) Animated Experience

Ga68, a GNU Algol 68 Compiler