Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
trostaft
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
61.
▲
by
trostaft
2y ago
'tree of arrays' Ahh, that's just Jax's concept of pytrees. It was something that they invented to make it easier (this is how I view it, not complete) to pass complex objects to function but still be able to easily cons
62.
▲
by
trostaft
2y ago
IIRC jax's `scipy.sparse.linalg.bicgstab` does support multiple right hand sides. EDIT: Or rather, all the solvers under jax's `scipy.sparse.linalg` all support multiple right hand sides.
63.
▲
by
trostaft
2y ago
You must be mistaken on the elden ring multiplayer bit, I was not too long ago summoned as a blue phantom when playing on my steam deck.
64.
▲
by
trostaft
2y ago
Seems like the guy’s website got hugged?
65.
▲
by
trostaft
2y ago
I don't know if google is somehow tracking the surge of interest in this article due to this HN post, but this very blog post is like search result 5 for me when looking up "how does simultaneous multi threading work" (result
66.
▲
by
trostaft
2y ago
I'm following the progress of this package, I would not mind a more modern successor to LaTeX. I'm an academic, and I actually do not mind LaTeX. I find actually writing LaTeX to be fairly natural. But every time I come to the pro
67.
▲
by
trostaft
2y ago
Ah, I'd love to try this. But I have a hard cross-platform requirement (Windows/Linux/MacOS) and I can't seem to get this running in WSL. Will keep checking if that improves in the future.
68.
▲
by
trostaft
2y ago
Well, I probably should not have played this, given that I'm (seemingly) in my last year of my PhD. Back to paper writing...
69.
▲
by
trostaft
3y ago
Because I always used to be self-conscious about the typing sounds in-lecture, I used to take hand-notes in class and then immediately after TeX them. While that had the happy accident of reinforcing the class notes, I find that having note
70.
▲
by
trostaft
3y ago
Unless I've misunderstood the comparison, the lists example does not appear to be comparing apples to apples. Reproducing here: $ python -m timeit "list((1, 2, 'a'))" 5000000 loops, best of 5: 53 nsec per loop
71.
▲
by
trostaft
3y ago
Well it plays all ways. PC misses out on the console exclusives, consoles miss out on the PC exclusives, consoles themselves have different exclusives and don't interop. For people getting into gaming when they're young, a big inf
72.
▲
by
trostaft
3y ago
Taking a look at a paper I have that went up this month and another that went up before the dec cutoff on ar5iv, they look 90% OK! Figures with side-by-side plots and algorithm environments are the common culprit for being broken though. Pa
73.
▲
by
trostaft
3y ago
I think this is what the top level poster is talking about? But this was 18 years ago. Maybe there was a more recent version https://www.computerworld.com/article/2568748/harvard-reject...
74.
▲
by
trostaft
3y ago
I wish the FEM ecosystem was a little more mature in Julia (particularly weak forms, their derivatives, and assembly). I'm confident about rewriting the other pieces of my research, but it took me a long time to be happy with FEniCS,
75.
▲
by
trostaft
3y ago
Hey, diffrax is cool by the way. I've used it to do some automatically differentiate through some ODEs for optimal experimental design. Great stuff.
76.
▲
by
trostaft
3y ago
My advisor gave me this paper at the beginning of my PhD; useful advice especially when google scholar dumps heaps of papers onto you every so often. Though I'd largely ignore the time estimates. That varies deeply depending on the pap
77.
▲
by
trostaft
3y ago
Seems like the bibtex on the page is broken? Or might just be an extension of mine.
78.
▲
by
trostaft
3y ago
Both Yahtzee (ZP) and Sebastian (Cold Take), who both ran my favorite two Escapist publications, have resigned along with him. I sincerely hope that the new management at the Escapist actually have a plan here, for the sake of the remaining
79.
▲
by
trostaft
3y ago
I find that frustrating with how intel markets its desktop CPUs. Often I find performance enhancements directly turning off efficiency cores...
80.
▲
by
trostaft
3y ago
... implying that they aren't already? At-least speaking US, there are plenty of real-estate focused PACs.
81.
▲
by
trostaft
3y ago
> At best they generate calls to high performing code. It should be noted that this is usually sufficient. But particularly for earth scale problems it can often not be.
82.
▲
by
trostaft
3y ago
It's nice to have easier to use C++ GPU numerical computing libraries, but it seems like there's a good bit of work to go in terms of feature set. Taking CuPy as an example, they've nearly entirely mirrored the enormous featu
83.
▲
by
trostaft
3y ago
Perhaps before killing someone with a comment, you should provide examples to back up your vitriol? The guidelines were reposted a mere four days ago... The writing felt fine to me, if a bit terse.
84.
▲
by
trostaft
3y ago
CUDA being feature complete and having libraries & frameworks optimized and immediately deploy-able for it is all the difference. I don't like much of their business/marketing practices for their gaming GPU division, but their
85.
▲
by
trostaft
3y ago
The Wikipedia article is pretty good, but tldr: "the overall performance improvement gained by optimizing a single part of a system is limited by the fraction of time that the improved part is actually used." https://en
86.
▲
by
trostaft
3y ago
> Speaking of telecom regulation, whatever happened to net neutrality? There was a good verge article on this https://www.theverge.com/2023/7/20/23800161/gigi-sohn-fcc-no...
87.
▲
by
trostaft
3y ago
Yes, to (grossly) summarize the conclusion of the paper: minres can outperform CG on the backwards error ||r_k|| / ||x_k|| whereas CG can outperform minres on the absolute error ||x^* - x_k|| and the energy norm ||x^* - x_k||_A.
88.
▲
by
trostaft
3y ago
I remember being handed this back when I was taking numerical analysis for the first time. It's an old document, but still useful. IMO the critical pieces of CG that make it a favorable choice for many problems in scientific computing
89.
▲
by
trostaft
3y ago
The particular pain point of multiprocessing in python for me has been the limitations of the serializer. To that end, multiprocess, the replacement by the dill team, has been useful as a drop in replacement, but I'm still looking for
90.
▲
by
trostaft
3y ago
It's a subjective question. Three that struck me recently were - Cyberpunk 2077 with its new path traced lighting option - Plague Tale Requiem (in general) - The high action sequences and cut-scenes of FFXVI These games have their flaw
More ›