bhouston

6 ms·

1.
Nvidia Launches RTX 5500 Pro Blackwell Desktop Edition with 84GB GDDR7 RAM(tomshardware.com)
6 points by bhouston 2d ago|4 comments
2.
WebGPU-Bench: A WebGPU Microbenchmark(ben3d.ca)
2 points by bhouston 2d ago|1 comments
3.
by bhouston 3d ago
I took it from here and I do see false transparency: https://superspl.at/scene/56155c3f There is even a link in the ThreeJS example preamble to this scene. I didn’t realize there was multiple lion splats of the same s
4.
by bhouston 3d ago
Because the benchmarks are stilling going up fairly rapidly. There are rumors of RSI. We are not at the top definitely. And then the following trend will be humanoids and similar and they are barely getting started.
5.
by bhouston 3d ago
It is an artifact of the model I believe. I did notice that as well. Gaussian splat models are fairly imperfect on average.
6.
by bhouston 3d ago
Demo here: https://threejs.org/examples/?q=gau#webgpu_gaussian_splat
7.
by bhouston 3d ago
> No, it's a high risk gamble. Even in a recession I think the shift towards AI would just accelerate since it is usually cheaper than humans. The real risk I feel if what if AI is too successful and there is a lack of human demand
8.
by bhouston 3d ago
It is a bit circular but it also a growing real market. NVIDIA is accelerating the market and also making money. It is a smart strategy on their part. We are not at the top of this trend yet so I do not see this as over investment. I remem
9.
by bhouston 4d ago
We are nowhere near a wall. Have you seen the progress of these models on benchmarks?
10.
by bhouston 4d ago
CEO trying to avoid government regulation in his industry says what?
11.
by bhouston 5d ago
The Houthis created a fake audio of a major Yemeni commander telling his troops to retreat which was subsequently amplified on Twitter/X, telegram, etc. Apparently, this helped the Houthis advance quickly as the opposition forces were
12.
by bhouston 5d ago
I think the key AI improvement is that these chips double the memory bandwidth and this has been the main limit for AI performance recently, became LLMs do not fit into cache.
13.
by bhouston 5d ago
I love the consistent CPU performance gains that Apple generates, always between between 15% to 30%. I know this is the standard since the Apple got on the ARM train, but for over a decade from 2008 to 2020 in the Intel world, there was ver
14.
by bhouston 5d ago
I think it depends on the MacBook Neo refresh frequency. When the MacBook Neo was released this year, it used the latest A-series chip (although non Pro version.). So when it is refreshed I suspect it will again use the latest A-series chip
15.
by bhouston 5d ago
Maybe you need a team of product managers/devs/designers then managing a set of apps? So while the ratio is roughly 1:1 for people to apps, you have that redundancy you want.
16.
by bhouston 6d ago
I think that Meta was just too early. I think that software team size per product (how ever you define that) are going to continue to shrink over the next 3 years. You will only need a single product designer/developer/manager pe
17.
by bhouston 6d ago
Yes, this is called RSI, e.g. recursive self-improvement. It is the current stage of things and it is part of a hard takeoff.
18.
Neural Texture Compression for Three.js(three-ntc.ben3d.ca)
4 points by bhouston 7d ago|0 comments
19.
How to Use Three.js's New Native Gaussian Splats(ben3d.ca)
28 points by bhouston 7d ago|10 comments
20.
Urban congestion relief experiments through routing-app interventions(nature.com)
2 points by bhouston 8d ago|0 comments
21.
Public WebGPU Benchmarking powered by vGPU(web3dsurvey.com)
2 points by bhouston 8d ago|0 comments
22.
by bhouston 8d ago
What happens to real fluid in this particular cases? If the singularity is in the physical space? Is this just a result of ignoring things like friction and energy dissipation via heat, etc?
23.
by bhouston 8d ago
Is this how politics works in the US? You hire the daughter of a prominent politician to get ahead?
24.
OpenAI hires Chuck Schumer's daughter as Big Tech cozies up to Democrats(nypost.com)
4 points by bhouston 8d ago|3 comments
25.
by bhouston 8d ago
> Ignoring the drama But the drama here is a little important. Stealing the millennium prize for N-S is sort of a big deal, especially to those who had been working on it for the last few years.
26.
Ant Colonies in the Machine(ben3d.ca)
1 points by bhouston 12d ago|0 comments
27.
by bhouston 12d ago
> I'm almost afraid to ask Claude if he could create a distributed LLM. Or you just add a lot of randomness to a bunch of small semi-smart LLMs. If you have enough of them, you basically are doing the "infinite monkeys" p
28.
by bhouston 12d ago
> Yes, ants that must be run on couch sized hardware drawing kilowatts continuously and generating text traces and CLI logs by the MB. I wrote this recently, basically mini lls that can run in any browser that has WebGPU support and ~4GB
29.
by bhouston 12d ago
I am starting to get the idea that AI feels like ants or weeds or mold. You simply can not get rid of it once you get an infestation. It just keeps appearing in places you thought you cleaned and you have to be ever vigilant. Right now gi
30.
by bhouston 13d ago
Okay. I will give a raw WebGPU implementation a try and see if it is smaller and more efficient.

More ›