Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
erwincoumans
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
31.
▲
by
erwincoumans
1y ago
I was assuming the context of robot learning (IsaacLab), where Newton Physics will eventually replace PhysX. Newton Physics doesn't target games or other areas.
32.
▲
by
erwincoumans
1y ago
The primary use case of Newton Physics is reinforcement learning, with 1000s of similar environments. Even if each environment would have sequential actions, you run many envs in parallel.
33.
▲
by
erwincoumans
1y ago
Well, MuJoCo initially used JAX for GPU (MJX) and MuJoCo Warp replaces MJX with better performance.
34.
▲
by
erwincoumans
1y ago
Agreed on most, and naming is terrible. Note that at run-time Python is out-of-the loop, since Newton Physics records a CUDA graph, and executes it, so performance is not impacted (aside from startup JIT time for modified kernels). I'd
35.
▲
by
erwincoumans
1y ago
This will eventually replace PhysX, some of its developers are working on Newton Physics. Newton Physics has multiple solvers, including MuJoCo-Warp and is easier to customize and extend.
36.
▲
by
erwincoumans
1y ago
iOS (iPhone, iPad) UI is typically smooth and fast though. If only car navigation and UI could be as responsive.
37.
▲
by
erwincoumans
1y ago
Agreed. However, according to the author LLM's mostly produce crap, and he doesn't seem to be able to imagine (or want?) that to improve (beyond crap/hallucination and become very useful to many).
38.
▲
by
erwincoumans
1y ago
A quick free way to access TPUs is through https://colab.research.google.com , Runtime / Change Runtime Type / v2-8 TPU
39.
▲
by
erwincoumans
1y ago
Those were fun times. A friend had a Power Cartridge and I reverse engineered it and re-implemented its features in the programmable Expert Express cartridge, using assembly language. It could also save/restore games, turbo loader and
40.
▲
by
erwincoumans
1y ago
that's a pretty WebGL fluid demo, it runs even well on iPhone.
41.
▲
by
erwincoumans
1y ago
Very nice to see. Maybe nice in VR, streaming to Quest 3/Apple Vision Pro (OpenVR/WebXR?)
42.
▲
by
erwincoumans
1y ago
MuJoCo, Drake, Pinocchio (and other simulators) are still improving (adding more accurate collision detection, better solvers etc).
43.
▲
by
erwincoumans
2y ago
Windows backward compatibility counter example: Paint Shop Pro 7.04 still works great here after almost a quarter century.
44.
▲
by
erwincoumans
2y ago
Thank you Macan! Asahi is a heroic effort, but way larger than a hobby. To be sustainable, it would require full buy-in from Linus and kernel maintainers indeed. Hope you enjoy well deserved better hobbies and family time.
45.
▲
by
erwincoumans
2y ago
Same here. >>contemporary C++30 can express the ideas embodied in such old-style code far simpler IMO, newer C++ versions are becoming more complex (too many ways to do the same thing), less readable (prefer explicit types over '
46.
▲
by
erwincoumans
2y ago
Except for the drifting joycons problems. We had to replace many. Hope Switch2 fixes that drift.
47.
▲
by
erwincoumans
2y ago
That's fun, it worked for me as well. I could spot the difference in all images, including the 'impossible mode'.
48.
▲
by
erwincoumans
2y ago
Like this? https://www.youtube.com/watch?v=YlDLwTsRHtU Ghost Robotics has been working with the US Army putting weapons on their robot quadruped for several years now. Unitree is ahead of the game with their hardware+softwa
49.
▲
by
erwincoumans
2y ago
It is a nice physics engine, it uses Taichi ( https://github.com/taichi-dev/taichi ) to compile Python code to CUDA/GPU (similar to what Warp Sim does, https://github.com/NVIDIA/warp )
50.
▲
by
erwincoumans
2y ago
You can copy C64 Basic to its underlying RAM (starting at $A000) and disable the ROM. Then it is mutable. https://www.lemon64.com/forum/viewtopic.php?t=69655
51.
▲
by
erwincoumans
2y ago
Thanks for sharing such an insightful article of complex material. Some follow-up showing how this helps optimization (gradient descent, Newton solvers?) would be great. >> For convenience of notation, we’ll drop the explicit parametr
52.
▲
by
erwincoumans
2y ago
Beautiful fragility captured.
53.
▲
by
erwincoumans
2y ago
If you like this, there is also the 500 lines of C++ TinyRenderer, loading Obj files, texture mapping, clipping, and vertex/pixel shaders on CPU only: https://github.com/ssloy/tinyrenderer
54.
▲
by
erwincoumans
2y ago
Thanks! From reading those older comments, the actual process is much more chaotic and iterative, closer to how Hayao Miyazaki creates his movies. See https://news.ycombinator.com/item?id=30275329
55.
▲
by
erwincoumans
2y ago
Warp outputs its intermediate GPU CUDA or CPU C++ files that can be compiled and linked into a binary. Here is an old example of mine calling Warp kernels from C++: https://github.com/erwincoumans/warp_cpp
56.
▲
by
erwincoumans
2y ago
Nice! Reminds me of the Op-1 Tombola sequencer: https://youtube.com/watch?v=SHoDUCAd4-I&pp=ygUMb3AtMSB0b21ib... that triggered midi notes on collision
57.
▲
by
erwincoumans
2y ago
Not mainstream but the game Teardown uses voxels, rendered on gpu, first screenspace , later voxel space: https://blog.voxagon.se/2018/10/17/from-screen-space-to-voxe... More recently Voxlands also renders
58.
▲
by
erwincoumans
2y ago
He mentioned the game Teardown, that allows destruction and physics on its voxels.
59.
▲
by
erwincoumans
2y ago
Interesting reading. I wonder if the author considered connecting the 5 volt line to a long lasting battery (with regulator) ? (and disconnecting the capacitor)
60.
▲
by
erwincoumans
2y ago
Sonic 10G rocks. I use Asus RT-AX89X router, with 2 10G ports, one input from Sonic, one output to 10G switches (Ubiquiti Networks UniFi Switch Flex XG 4-Port 10G).
More ›