krapht

6 ms·

31.
by krapht 8mo ago
Why not have a megatopic for things that don't need their own topic? Topics are necessary when you start having a huge Zulip server, 100+ people. There's so much noise --- dividing things by channel is too coarse. I participate in
32.
by krapht 8mo ago
I've seen some small firms crash and burn too, though. The problem is small firms are easy-come, easy-go; they don't have enough reputation at stake. Not sure what a good solution is.
33.
by krapht 9mo ago
Yeah, I know three - one transitioned to teaching, another to being a paramedic, and the last to social work. On the other side I also know a teacher who switched to cyber security for the money after he started a family. You have to know y
34.
by krapht 9mo ago
What you suggested has been done before - you might find a review of the literature fun if this sort of thing interests you, even if academic papers are pretty dry reading normally.
35.
by krapht 10mo ago
It wasn't labor costs that made the Roomba twice a Roborock, but manufacturing costs. Roomba teardowns basically show that iRobot was just really bad at cost-optimizing their vacuums for mass production.
36.
by krapht 10mo ago
There's no way to get through the harder courses in the program on 1 hour a day. And you're not getting value from the degree if you aren't pushing yourself to take those hard courses, unless you just need the diploma.
37.
by krapht 10mo ago
This is a standard which few kernels will ever meet. I'd say requiring a numerical proof is the same as requiring no proof at all - because it won't ever happen unless you're validating silicon or something equally expensive.
38.
by krapht 11mo ago
Why can't delayed spending be categorized as a claim to future goods (savings)?
39.
by krapht 11mo ago
C++: "look at what others must do to mimic a fraction of my power" This is cute, but also I'm baffled as to why you would want to use macros to emulate c++. Nothing is stopping you from writing c-like c++ if that's what
40.
by krapht 11mo ago
Good luck displacing MATLAB, it's great there's an OSS alternative here.
41.
by krapht 11mo ago
> The benefit of keyboard-driven programs like Vim is that you're trading an initial learning curve for a vastly more efficient experience once the learning is done+. I have never been rate-limited by my keyboard input speed. I have
42.
by krapht 11mo ago
Agree on shorter, disagree on simpler. The hard part of understanding GPU code is knowing the reasons why algorithms are the way they are. For example, why we do a split-k decomposition when doing a matrix multiplication, or why are we load
43.
by krapht 11mo ago
Funny, I feel the same way about Triton. Performant Triton looks like CUDA (but with tiles!) except it's ten times harder to debug since it doesn't have the tooling NVIDIA provides. If I had to run on AMD I'd rather deal with
44.
by krapht 11mo ago
This is cool. Not personally going to switch because I want to stay with the official implementation, but I appreciate the effort involved in porting libraries.
45.
by krapht 11mo ago
Citation needed.
46.
by krapht 11mo ago
Anaconda is a good example of this.
47.
by krapht 1y ago
Have you used Google wuffs?
48.
by krapht 1y ago
There is no way to write inline SASS (assembly equivalent) for CUDA code. You can inline PTX, but PTX is a high level bytecode designed to be portable.
49.
by krapht 1y ago
An excellent book for fundamentals. Still haven't found a good textbook that covers the next level, that takes you from a student to competent practitioner. Advanced knowledge that I've picked up in this field has been from cowork
50.
by krapht 1y ago
I'll believe it when I see it reflected in applicant resumes. (east coast tech firm)
51.
by krapht 1y ago
I've seen media reports from China about elevated lung cancer rates in non-smoking women. Just as scary, IMO. That being said, Chinese cooking makes much greater usage of stir-frying, and it's well known that most residential ven
52.
by krapht 1y ago
My opthalmologist told me I had enough corneal thickness to get PRK done again if I wanted.
53.
by krapht 1y ago
You couldn't in the 50s either. You just made do.
54.
by krapht 1y ago
Ahh, vibe coding.
55.
by krapht 1y ago
Nothing, but what stops you from using DEFLATE64? Installing new software has a real time and hassle cost, and how much time are you actually saving over the long run? It depends on your usage patterns.
56.
by krapht 1y ago
Variable length parallelism is hard. You can go to highload.fun (SIMD competition site) for problems that are only parallelized after significant effort. Try problem #1, parsing numbers.
57.
by krapht 1y ago
What's the difference between this and registering pre and post commit hooks?
58.
by krapht 1y ago
Sure, but at that point, why not just Minecraft then?
59.
by krapht 1y ago
If this is the only issue, can't this be addressed by normalizing the post-processed data before scoring? (that is, if it really is just a matter of block ordering)
60.
by krapht 1y ago
Yes, and deep research was junk for the hard topics that I actually needed to sit down and research. Anything shallower I can usually reach by search engine use and scan; deep research saves me about 15-30 minutes for well-covered topics. F

More ›