bssrdf

6 ms·

1.
The End of Programming(pauldix.com)
14 points by bssrdf 1mo ago|1 comments
2.
by bssrdf 3mo ago
I found [ggml] is really easy, educational and fun to use. It is written in C and the backbone for llama.cpp. [ggml] https://github.com/ggml-org/ggml
3.
None of this is complicated. I'm just way too lazy to write my own CUDA code(twitter.com)
2 points by bssrdf 2y ago|0 comments
4.
by bssrdf 2y ago
I would like to recommend "Understanding UNIX/LINUX Programming: A Guide to Theory and Practice" for beginners. You can download source code I assembled https://github.com/bssrdf/UnderstandingUnixLinuxPro
5.
Train a Mnist VAE with C and CUDA(github.com)
54 points by bssrdf 2y ago|2 comments
6.
A curious case of O(N^2) behavior which should be O(N) (2023)(gist.github.com)
101 points by bssrdf 2y ago|44 comments
7.
Genesis: A generative and universal physics engine for robotics and beyond(genesis-embodied-ai.github.io)
83 points by bssrdf 2y ago|2 comments
8.
by bssrdf 2y ago
Indeed. Even a simple 'wc -l' type operation can benefit from SIMD [0] [0] https://x.com/cloud11665/status/1799534538873290993
9.
by bssrdf 3y ago
Yes, actually it is a doubly linked list.
10.
by bssrdf 3y ago
I don't think a PR is ready. My solution does not fully solve the problem. It requires more thoughts. Thanks.
11.
by bssrdf 3y ago
Good point, I have the same doubt. Even if it is sorted, you still cannot do binary search on it unless using some skip list ideas. In fact, the code right now is doing linear search to find the ID given a name. TBH, I don't fully unde
12.
by bssrdf 3y ago
I didn't submit the PR yet, but opened an issue which has not been picked up. Maybe most of the USD files used in Blender do not trigger the O(N^2) behavior.
13.
by bssrdf 3y ago
Just finished writing [0] about Blender's USD import issue. [0] - https://news.ycombinator.com/item?id=36947946
14.
A curious case of O(N^2) behavior which should be O(N) (2023)(gist.github.com)
27 points by bssrdf 3y ago|0 comments
15.
by bssrdf 5y ago
The key word: TUNING. Every climate model has tons of tunable parameters, many of which are not from first principles. They are empirical, observed some years ago over some remote locations, but now applied globally as the truth.
16.
by bssrdf 5y ago
I implemented radix sort in cpython [0] and did some simple benchmarking comparing to python' built-in sort (Timsort I believe). It is indeed faster. [0] https://github.com/bssrdf/RadixSortPy
17.
Some Lecture Notes for Algorithms(timroughgarden.org)
1 points by bssrdf 5y ago|0 comments
18.
Mythbusting Competitive Programming – You don't need to learn it(freecodecamp.org)
1 points by bssrdf 6y ago|2 comments
19.
by bssrdf 6y ago
This book <<The Nature of Computation>>[0] has the best explanation I have ever seen. [0]: https://www.amazon.com/Nature-Computation-Cristopher-Moore-d...
20.
A performance-portable atmospheric dynamical core in C++(github.com)
1 points by bssrdf 7y ago|0 comments
21.
Self-driving AI task running on a 0.9GHz CPU processor(youtube.com)
3 points by bssrdf 8y ago|1 comments
22.
Furloughed workers trying to make ends meet(washingtonpost.com)
2 points by bssrdf 8y ago|0 comments
23.
IBM announces high-resolution global weather forecast model(washingtonpost.com)
1 points by bssrdf 8y ago|0 comments
24.
Science insurgents plot a climate model driven by artificial intelligence(sciencemag.org)
2 points by bssrdf 8y ago|0 comments
25.
Reflections on Three Years of Reading Knuth(commandlinefanatic.com)
20 points by bssrdf 8y ago|2 comments
26.
Deep Learning – From Basics to Practice(github.com)
3 points by bssrdf 9y ago|0 comments
27.
Their eyes are on D.C.: Capital area said to have good shot at Amazon HQ2(washingtonpost.com)
7 points by bssrdf 9y ago|2 comments
28.
I do like CFD(cfdbooks.com)
23 points by bssrdf 9y ago|5 comments
29.
Goodbye to Twitter (and all that)(pharr.org)
1 points by bssrdf 9y ago|0 comments
30.
Reprogramming my career(science.sciencemag.org)
3 points by bssrdf 9y ago|0 comments

More ›