brrrrrm

36 ms·

1. 421.▲
   by brrrrrm 5y ago

   great writeup! For those interested, I was only able to get the demos running in Chrome Canary with "Unsafe WebGPU" enabled.
2. 422.▲
   by brrrrrm 5y ago

   I've found shared tmux/screen sessions are ideal. + each person can have their own font/resolution + it doesn't dominate the entire screen, allowing each person to keep their own notes/etc on the side + at any point
3. 423.▲
   by brrrrrm 5y ago

   This is a demo of tensor programming in real time to manipulate a webcam feed. There are a couple of examples that can be loaded (top left) such as edge detection and color manipulation. The code is JIT compiled to WebAssembly and all the
4. 424.▲
   Show HN: Realtime programmatic image manipulation in the browser(loop-tool.glitch.me)

   3 points by brrrrrm 5y ago|2 comments
5. 425.▲
   The most modern JavaScript I know(jott.live)

   1 points by brrrrrm 5y ago|0 comments
6. 426.▲
   by brrrrrm 5y ago

   I couldn’t interact with the sidebar on mobile. It kept either reloading the page or scrolling to the top
7. 427.▲
   by brrrrrm 5y ago

   Don’t compilers write code for you?
8. 428.▲
   Show HN: Loop_tool.js demo, a JIT WASM compiler for linear algebra (and ML)(loop-tool.glitch.me)

   3 points by brrrrrm 5y ago|0 comments
9. 429.▲
   by brrrrrm 5y ago

   ^ this. Ref: https://webassembly.github.io/spec/core/syntax/types.html#sy...
10. 430.▲
    Show HN: Loop_tool's Interactive Loop Optimization(jott.live)

    2 points by brrrrrm 5y ago|0 comments
11. 431.▲
    by brrrrrm 5y ago

    I just finished adding a basic WASM[1] backend + basic JavaScript frontend[2]. I'm in the process of adding in-browser optimization[3] and will hope to have a demo some time this week! [1] https://github.com/facebookre
12. 432.▲
    by brrrrrm 5y ago

    I was curious about the same thing, here's a list I found: https://webgl2fundamentals.org/webgl/lessons/webgl2-whats-ne... I guess a big one outside of graphics is integer textures and bit manipulations. The
13. 433.▲
    by brrrrrm 5y ago

    video rendering wasn't working in chrome/firefox a second ago, I just fixed it!
14. 434.▲
    Show HN: Interactive Loop Optimization(jott.live)

    10 points by brrrrrm 5y ago|3 comments
15. 435.▲
    by brrrrrm 5y ago

    re: leading comparison, that makes sense. I recommend labeling it a bit better: JAX on TPU vs NumPy on CPU
16. 436.▲
    by brrrrrm 5y ago

    To clarify, I’m not in any way involved with the paper described in the article. Just interested in the space :)
17. 437.▲
    by brrrrrm 5y ago

    I'd be curious to read up on the parts of XLA design that were inspired by Halide if you have any links. I haven't familiarized myself with XLA/MLIR source code much, but I was under the impression it took a more conventiona
18. 438.▲
    by brrrrrm 5y ago

    You're correct that low level details very much matter in the HPC space. The types of optimizations described in this paper are exactly that! To elucidate, check out Halide's cool visualizations/documentation (which this pa
19. 439.▲
    by brrrrrm 5y ago

    It'd be quite a boring world if observations of the past could reliably predict the future.
20. 440.▲
    by brrrrrm 5y ago

    a pure functional language is trivially differentiated, no? View a program as a data flow graph and then append the differentiation data flow graph. How this is executed (reverse-mode vs forward-mode vs some exotic mix) is just a leaked a
21. 441.▲
    by brrrrrm 5y ago

    for the type of computation described in this article, Rust is considered slow. I recently stumbled across a site to track it's progress in changing that: https://www.arewelearningyet.com
22. 442.▲
    TV Pickup(en.wikipedia.org)

    2 points by brrrrrm 5y ago|0 comments
23. 443.▲
    by brrrrrm 5y ago

    Any benchmarks?
24. 444.▲
    Ask HN: How do I learn to write good modern Rust code?

    18 points by brrrrrm 5y ago|2 comments
25. 445.▲
    by brrrrrm 5y ago

    on an M1 I think TF.js is within 2x of native XNNPACK on a single core. Since aarch64 simd width is also 128, it works out well.
26. 446.▲
    by brrrrrm 5y ago

    Isn't this readily solved by a user convention such as only updating one variable each cell? Why limit users based on mistakes they haven't made yet?
27. 447.▲
    by brrrrrm 5y ago

    The emphasis of this project seems to be text based coding and visual description of dependencies at the cell level. Visual programming, at least according to wikipedia, refers to "manipulating program elements graphically rather than
28. 448.▲
    by brrrrrm 5y ago

    I like that this is basically Jupyter/Colab extended with an explicit cell dependence graph. However, I think this needs some serious brand work to distinguish it from "visual programming," which multiple commenters have alre
29. 449.▲
    Show HN: Emojraw – Draw with an Emoji Palette(emojraw.glitch.me)

    4 points by brrrrrm 5y ago|1 comments
30. 450.▲
    by brrrrrm 5y ago

    Beyond debugging, it’s a useful idea for tweaking the trade off between compile time and resultant performance. Especially true with JIT or runtime compilation.