jacobgorm

3 ms·

1.
Neso: Metal MSL and D3D12 HLSL Back End for Triton(github.com)
1 points by jacobgorm 8d ago|1 comments
2.
by jacobgorm 8d ago
Neso is an ahead-of-time compiler backend for Triton. It lowers Triton TTIR/TTGIR into MSL or HLSL shader code, to allow modern LLM compute kernels such as Flash Attention II to run on GPUs in consumer devices such as Macs and Windows
3.
by jacobgorm 9d ago
I can program all my non-CUDA GPUs use completely open non-proprietary toolchains. And if that ceases to be the case on one platform I can switch platforms without having to rewrite all my code.
4.
by jacobgorm 9d ago
If that business makes it money selling a cross-platform AI inference engine, as was the case for my previous startup, it is bad business to tie yourself to single platform. I managed to build a single code base that would support CPUs, Ope
5.
by jacobgorm 9d ago
It is currently faster than the stock Candle / MPS shaders it replaces on MacOS/ARM64, and IIRC a bit slower than OpenVINO/CPU on my old Windows laptop, where I never got OpenVINO/GPU to compute correctly. Candle didn&#x
6.
by jacobgorm 9d ago
As it happens, I just got my employer's permission to release as open source a Triton back-end for Metal and D3D12 GPUs here: https://github.com/dropbox/neso . As an example of you how can use it to deploy real mo
7.
by jacobgorm 9d ago
What reasons would you have to prefer Rust over C for compute kernels? I am a great fan of Rust, but I don't see any benefit for kernels, due to their relatively simple nature.
8.
by jacobgorm 9d ago
For learning that may be a fine approach, but CUDA (in C++) really tries to hide what is going on behind the scenes, which is roughly: 1) code gets split between a host part that goes through your normal compiler, and a device part that goe
9.
by jacobgorm 9d ago
Because I know it is not technically crap, a lot of competent people worked on it, most with good intentions. I suppose it is better described as a cleverly designed Trojan horse than can infect your software and make that software become c
10.
by jacobgorm 9d ago
I haven't felt this popular since then 1990s when I was opposing Visual J++ and IIS.
11.
by jacobgorm 9d ago
You can call their APIs without needing to compile your code with a proprietary compiler or adopt a bastardized version of C++.
12.
by jacobgorm 10d ago
No. CUDA allows you to write all the code in a single file, and uses a preprocessor to split it back out and pass it through separate compilers, one for host and one for device.
13.
by jacobgorm 10d ago
CUDA is not an API, CUDA is a language, so you cannot make that comparison.
14.
by jacobgorm 10d 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.
15.
by jacobgorm 10d 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 progr
16.
by jacobgorm 13d ago
I invented VM live migration to solve the problems with traditional process migration, and my work was inspired by both Sprite and MOSIX. I believe that VM migration is in wide use today, especially at hyperscalers. Mike Nelson, who worked
17.
by jacobgorm 1mo ago
I've been on Mastodon for a few years, and though it has less drama than Twitter/X, it does seem to have reached critical mass of users and does cover most of my needs for random gossip. I am surprised that that nobody has mention
18.
by jacobgorm 1mo ago
You were lucky to get out alive.
19.
by jacobgorm 1mo ago
You can always hide in your own country, as long as you or the people who protect you are able to stay in power, but at least you cannot use the rest of the world as playground or shopping center. Sucks to be a rich kleptocrat and not be ab
20.
by jacobgorm 1mo ago
There is no memory overhead, in that everything resides on disk/in the sqlite database, and is read from there on demand. That said, sqlite for blob storage is a bit inefficient, so the next version will use tightly packed structs on d
21.
by jacobgorm 1mo ago
Here is a project with similar goals (I am the main author), but using late interaction semantic search: https://github.com/dropbox/witchcraft
22.
by jacobgorm 3mo ago
I just uninstalled it FWIW.
23.
by jacobgorm 4mo ago
Seems very likely, related: https://electrek.co/2026/05/31/a-spacex-tesla-merger-could-t...
24.
by jacobgorm 4mo ago
Make sure you watch the Big Blue, in the original French (not US, has the wrong soundtrack) version, but with the original(!) English speech, not the dubbed-to-French version.
25.
by jacobgorm 4mo ago
Smoke free perhaps but not nicotine free by far. I live in Denmark and have a teenage son, he tells me that 90% of girls at his year are hooked on illegally imported and unregulated vapes, much stronger in nicotine than cigarettes, and in s
26.
by jacobgorm 5mo ago
It is not that long ago that smartphones would die from moisture exposure if you left them in the bathroom while taking a shower. I had a girlfriend around 2001 who spent all her savings on a shiny new Nokia 8250, got drunk and barfed on he
27.
by jacobgorm 5mo ago
In Denmark, wind mills were initially quite popular, because locals owned them and benefited. The iconic wind farm Middelgrunden on the waters outside of Copenhagen is 50% owned by a co-op. https://www.middelgrunden.dk/
28.
by jacobgorm 5mo ago
Good point, an int3 is not going to be faster than a syscall, and if they implement the sandboxing policy in guest userspace is seems it would be quite easy to disable.
29.
by jacobgorm 5mo ago
I think the point here is optimizing for the common case, the untrusted code is still running inside a VM, so you can still trap malicious or corner cases using a more heavy-handed method. The blog post does mention "self-healing"
30.
Show HN: Witchcraft and Pickbrain – fast multi-vector semantic search in Rust(github.com)
6 points by jacobgorm 5mo ago|0 comments

More ›