frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Nvidia announces native GPU programming in Rust

https://developer.nvidia.com/blog/introducing-cuda-rust-two-tracks-for-writing-gpu-kernels/
75•nonmaskable•11h ago

Comments

the__alchemist•52m ago
I'm looking forward to trying these when they stabilize! I currently use WGPU for graphics, and cudarc for CUDA.

Note: Cuda-oxide is similar to Cudarc's host component, but uses a rust-style kernel dialect. Advantage: Share structs between host and device. Disadvantage: Trading standard Cuda kernels for a new, WIP dialect.

I haven't tried the tile API yet; looking forward to it.

The last time I checked, Cuda Oxide was Linux only, and required Async; these are why I haven't tried it yet.

embedding-shape•37m ago
cudarc been great for me, because it's easy to look up existing examples and references, and it maps 1-to-1 with what I see. I'm already having a tough time with CUDA itself, a dialect of it makes a tad harder to rely on previous work.

Seems more ergonomic in general though, both approaches they share, compared to cudarc, and less build infrastructure and fiddling with environments, which is great.

dllu•50m ago
Since NVIDIA owns huggingface now and huggingface has the excellent Candle [1] crate for inference on Rust, this seems like a good step towards nice native Rust kernels.

[1] https://github.com/huggingface/candle

jacobgorm•17m ago
Nobody cares if kernels are written in Rust. Kernels were meant to be written in C, but if you want to go more high-level try Triton or a similar DSL that nicely abstract tile sizes etc.
claiir•45m ago
> The launch is checked rather than trusted.

Damn even Nvidia is putting out fully Claude-written articles.

manyatoms•13m ago
not to worry, they have an 'AI generated summary' box too
mahboi•4m ago
Thanks, saved me a few minutes
bayindirh•2m ago
That's actually a magnificent observation. This is not only an indication of a keen eye, but a trained brilliant mind as well.
rvz•42m ago
First of all, this is a pre-1.0 release that requires a nightly Rust compiler (if you choose the SIMT track with cuda-oxide) so that one is going to be unstable software.

Secondly, When an issue occurs with a kernel or you want to write your own custom kernel in Rust, now we need to diagnose if the problem came from either cuda-oxide (SIMT), Rust's side, CUDA or Tile (If you decide to choose the Tile track).

Another dependency into the list and course everything is open source except CUDA itself. So any issue that happens on the CUDA level, you are forced to wait for them to fix it.

LarsDu88•29m ago
In this age of LLM written everything which has softly killed my motivation for learning Rust somewhat, this has revived my interest if not only for the fact the LLMs haven't yet been trained on this yet!
impulser_•23m ago
LLM don't need to be trained in a library to use it well. It's just Rust which they know well.
w4yai•5m ago
And what prevent you exactly ?

There were humans far superior than you for writting Rust before LLM, now there's a LLM. The only difference is price and time execution.

You get an awesome teacher (LLM) ready to answer all your questions about Rust.

And you still find excuses not to learn it ?

At some point, just realize you've been lazy to learn it and LLMs are just an excuse.

nicebyte•26m ago
what this article tells me is that no one at Nvidia actually cares about this project whatsoever. otherwise, they would have had a person actually write the announcement.
jacobgorm•19m ago
I strongly dislike CUDA. Once you have allowed that proprietary cr*p into your C++ codebase, it is very hard to get rid, and you end up with code that is either tied to a single vendor or an #ifdef hell, probably both.

The best way to program GPUs is face up to the reality that they are not the same machine as the CPU, write your kernels in separate files, and launch them manually, like in Metal, OpenCL, and D3D12, etc. These days we even have DSLs like Triton that make kernel writing much more ergonomic than anything you would hope to achieve in Rust.

bigyabai•13m ago
Is this satire? D3D12 and Metal aren't any less proprietary than CUDA.
pavon•11m ago
> The best way to program GPUs is face up to the reality that they are not the same machine as the CPU, write your kernels in separate files, and launch them manually

Isn't that how CUDA code is normally written?

melodyogonna•9m ago
You could also use Mojo, one language for all targets.
carefree-bob•8m ago
I began to lose interest after the acquisition. Have you been following along, are they still going to open source it?
YuechenLi•3m ago
manyatoms•12m ago
How does this compare to vectorware? (https://www.vectorware.com/blog/)
I thought they already did and released the compiler source code under Apache 2.0.

Training a 4B model to produce 81% faster query plans than Postgres

https://rohanbansal.com/qorl
317•polyphilz•4h ago•60 comments

Breaking the 1.58-bit Barrier for Ternary LLMs

https://arxiv.org/abs/2609.16338
86•matt_d•2h ago•6 comments

Nvidia announces native GPU programming in Rust

https://developer.nvidia.com/blog/introducing-cuda-rust-two-tracks-for-writing-gpu-kernels/
75•nonmaskable•11h ago•21 comments

Xiaomi Mimo 2.6 live post-training dashboard

https://mimo.xiaomi.com/rl/
174•krackers•3h ago•48 comments

OpenSpec – A lightweight and configurable AI spec framework

https://openspec.dev/
5•etoxin•7m ago•0 comments

Small programming tricks

https://will-keleher.com/posts/small-programming-tricks-matter/
344•signa11•7h ago•171 comments

Backups Aren't Simple

https://filipovski.net/2026/09/16/backups-arent-simple.html
19•afilipovski•2h ago•4 comments

Reversing Factorio's RNG

https://gegell.github.io/posts/factorio-rng/
98•jheitmann•4d ago•11 comments

Performance Improvements in .NET 11

https://devblogs.microsoft.com/dotnet/performance-improvements-in-net-11/
112•soheilpro•1d ago•14 comments

AWS says it can't restore some data from mideast facilities struck by Iran

https://www.wsj.com/world/middle-east/aws-says-it-cant-restore-some-data-from-mideast-facilities-...
149•berkeleyjunk•1d ago•120 comments

The engineering behind the US Strategic Petroleum Reserve

https://johnjwang.com/post/2026/09/15/engineering-behind-us-strategic-petroleum-reserve
22•johnjwang•1d ago•1 comments

Accurate Models of AMD Matrix Cores

https://arxiv.org/abs/2609.14845
48•matt_d•4h ago•6 comments

Japan's book scene is moving from bookstores to libraries

https://untranslatedjp.substack.com/p/japans-book-scene-is-quietly-moving
66•herbertl•4d ago•19 comments

How good are frontier models at physics?

https://arxiv.org/abs/2609.13009
54•qt31415926•3h ago•25 comments

Dream-RSI: Recursive Self-Improvement through Evolving Worlds

https://arxiv.org/abs/2609.14858
172•bananaflag•9h ago•49 comments

Mistral X Mozilla: Private, Multilingual AI Browsing

https://mistral.ai/news/mistral-x-mozilla/
511•vertigoruntime•15h ago•182 comments

Anatomy of a Texture

https://agentlien.github.io/texture/
65•Agentlien•8h ago•11 comments

Show HN: An e-ink frame that hears birds and draws them as 1800s illustrations

https://github.com/arnegiacomo/fugleramme
2040•arnemunthekaas•1d ago•236 comments

Reverse-engineered Jev-like model

https://github.com/vinnylarouge/jevlike
39•rochansinha•4h ago•6 comments

Anecdotally, programmers dislike "reduce"

https://evanhahn.com/posts/2026-09-13-programmers-dislike-reduce/
66•vinhnx•2d ago•112 comments

WalShadow: Sub-second Postgres replication to ClickHouse from physical WAL

https://clickhouse.com/blog/introducing-walshadow
33•spathak•5d ago•5 comments

Why Does the Universe Expand?

https://cosmicave.org/2026/09/15/why-does-the-universe-expand/
7•the__alchemist•3h ago•1 comments

I replaced my brown-noise browser tab with a menu bar app

https://oldmanrahul.com/2026/09/14/hush/
7•oldmanrahul•10h ago•0 comments

Vectorized and performance-portable Quicksort (2022)

https://opensource.googleblog.com/2022/06/Vectorized%20and%20performance%20portable%20Quicksort.html
160•mococa•4h ago•24 comments

Hackers Got Inside a Flock Camera

https://www.wired.com/story/hackers-flock-camera-data-shows-how-system-works/
428•driverdan•9h ago•206 comments

The DeepMind Institute

https://institute.deepmind.com/
121•vertigoruntime•8h ago•41 comments

Tell the speakers that you liked their talks

https://ohhelloana.blog/tell-the-speakers/
263•whisper2020•1d ago•74 comments

Kyber (YC W23) Is Hiring a Forward Deployed Engineer

https://www.ycombinator.com/companies/kyber/jobs/eturrAR-forward-deployed-engineer
1•asontha•11h ago

Training Text-to-Image Models 3.6× Faster

https://www.linum.ai/field-notes/jit-ddt
20•schopra909•6h ago•1 comments

The Siberian Ice Maiden and the Scythian World

https://patrickwyman.substack.com/p/the-siberian-ice-maiden-and-the-scythian
45•NaOH•1d ago•4 comments