Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
nallana
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
1.
▲
by
nallana
6d ago
Also on the team here. Feel free to ask any questions — we’ll be monitoring this thread for the next few hrs! Curious what other’s views on DSLs are post-LLMs too
2.
▲
Show HN: Adding and Extending Integer Support for MATLAB Code in RunMat
(runmat.com)
4 points
by
nallana
19d ago
|
0 comments
3.
▲
by
nallana
5mo ago
They haven’t been all this time? Damn — what a time to be a robot
4.
▲
by
nallana
6mo ago
Why not a shared buffer? Serializing into JSON on this hot path should be entirely avoidable
5.
▲
by
nallana
9mo ago
Appreciate the comment actually! It's good feedback -- we weren't sure if it's mixing work/memes too much, and keeping our materials clean like engineering docs are probably a good way to go. We may edit it out of the po
6.
▲
by
nallana
9mo ago
It's open source / free. But yes, of course we want people to try it and get value from it.
7.
▲
by
nallana
9mo ago
Seeing in MATLAB code how ode45 is implemented != how the thing is running on the machine. That's a very small top slice. But okay -- as I mentioned, you're entitled to your views!
8.
▲
by
nallana
9mo ago
Shameless plug for RunMat (we wrote this blog article, also an open source alternative for MATLAB): https://runmat.org
9.
▲
by
nallana
9mo ago
Yep -- Octave was very helpful for me in school. Octave is not particularly fast. RunMat is very fast (orders of magnitude -- see benchmarks).
10.
▲
by
nallana
9mo ago
>> The engine is closed source. You cannot see how fft or ode45 are implemented under the hood. For high-stakes engineering, not being able to audit your tools is a risk. This is just a lie. Open matlab and you can inspect all the imp
11.
▲
by
nallana
9mo ago
If the rewrite is for performance, ideally the logic capture = the thing executing in production. Cases where a JIT running would conflict with requirements notwithstanding (e.g. HIL with strict requirements and whatnot)...
12.
▲
by
nallana
9mo ago
Yep! Makes sense. Though I think the cost of writing these toolboxes is lim --> 0. Will have a really solid rust inspired package manager soon, and a single #macro to expose a rust function in the RunMat script's namespace (= easy t
13.
▲
by
nallana
9mo ago
In Julia, you explicitly need to still reason about and select GPU drivers + manage residency of tensors; in RunMat we abstract that away, and just do it for you. You just write math, and we do an equivalent of a JIT to just figure out when
14.
▲
by
nallana
9mo ago
@mNovak -- super helpful note! Thank you! Author of RunMat (this project) here -- > The first thing they teach about performant Matlab code is that simple for-loops will tank performance. Yes! Since in RunMat we're building a comput
15.
▲
by
nallana
10mo ago
Thanks!! It was originally for Octave users whose scripts were running painfully slow. The goal was to keep the MATLAB frontend capture syntax, but run it fast. When we dug into why people were still using Octave, it was because it let th
16.
▲
Show HN: RunMat – runtime with auto CPU/GPU routing for dense math
(github.com)
21 points
by
nallana
10mo ago
|
4 comments
17.
▲
Choosing Rust for LLM-generated code
(runmat.org)
7 points
by
nallana
1y ago
|
4 comments
18.
▲
by
nallana
1y ago
a solid core, not the whole Matlab (they confound the language, compiler / runtime, an IDE, and a bunch of other things in the name / product that is MATLAB). This is a solid compiler + minimal runtime, with an architecture design
19.
▲
by
nallana
1y ago
Versus Octave: - Lots of language semantics are unsupported in Octave (like Classes), and it’s purely a slow line by line interpreter so it’s very slow. - Given the design / Cranelift IR translation here, RunMat can run natively on any
20.
▲
Show HN: RunMat – a V8 inspired Rust runtime for the landlocked Matlab language
(runmat.org)
15 points
by
nallana
1y ago
|
6 comments
21.
▲
Show HN: Using generative AI for real world engineering analysis
(dystr.com)
17 points
by
nallana
4y ago
|
1 comments
22.
▲
by
nallana
4y ago
Hi HN! Would love to share something that has been over a year in the making. We were originally building an environment to conduct engineering / scientific analysis in the cloud (lambda compute w/ visual outputs for figures and w
23.
▲
Show HN: A new way to write and publish research
(dystr.com)
39 points
by
nallana
4y ago
|
2 comments