51–60 of 177 posts
Something I'm personally excited about: colour e-ink displays. There are a few decent colour e-ink tablets on the market already, for surprisingly low prices ($400-500). So far the colour reproduction and refresh rate is still somewhat suboptimal, but development appears to be progressing rapidly. I predict that in a few years there will be little reason to get a conventional tablet over an e-ink one, at least for th…
We've all pretty much forgotten about AR/VR But Apple's going to be launching their headset next month
WebGPU. It’ll be live in Chromium stable this Thursday and will be the biggest step change for browser-based GPU since the original WebGL.
Mars Sample Return going through PDR. Intrinsic AI making a product announcement in two weeks. Deepmind collaborating with Robotis implementing deep RL on the robocup humanoids to play soccer at the next level.
like usual, basic php + mysql apps making money . they will never die
Now that the NFT and Meta clown cars have crashed, metaverse work is quietly plugging along. It's a niche, but it may be a Roblox/Fortnite sized niche.
Is anything actually exciting happening in the space, though?
Epic released Unreal Engine for Fortnite about one month ago.
Meta have some interesting stuff in their pipe.
I think that especially Roblox will shift how the younger generation view games, with more emphasis on social. It is their MMOs.
Earlier quoted context omitted.
Yes, it's definitely lagging the CUDA ecosystem. I did a write-up a few weeks back on the various approaches people are taking to do inference in the browser (WASM, WebGL, WebGPU): https://digest.browsertech.com/archive/browsertech-digest-th... Matt Rickard also had a short write-up recently: https://matt-rickard.com/implementing-llms-in-the-browser My take is that we probably won't see much training done on WebGPU,…
> But for inference and run-time computations, it could be very interesting to take a model trained with CUDA/PyTorch and export it (maybe with Apache TVM or tensorflow.js) into WebGPU that can run on end-user devices. In its current state, can you train on PyTorch, export to ONNX, load ONNX in JavaScript/WASM, then use it for WebGPU inference? I'm not trying to sound obsessed/married to ONNX, I just though it was "t…
I believe so. Onnxruntime very recently merged a WebGPU backend: https://news.ycombinator.com/item?id=35694553
You can also go directly from PyTorch to WebGPU with Apache TVM. (ONNX is also supported, but my understanding is that it's better to go direct). This is an example using an LLM trained with PyTorch (I think) and run in the browser: https://mlc.ai/web-llm/
One of the questions I've been asking is we've basically had physical servers -> VM -> cloud -> container orchestration on about a decade long cadence. So what's next--especially considering you can argue a lot of the low-level architectural things haven't really changed during that period. But, then, perhaps architectural details don't matter a lot at this point relative to things like AI/ML. Even WASM is mostly in…
The trendy (meaning both that it is hyped and that it follows the trajectory of the trend you identified) answer is serverless/scale-to-zero (Lambda, Cloud Run, etc.). That said, Lambda has been out for a while and seems to have avoided taking over the world. Generic vCPU-hours are so cheap now though that it isn't compelling from a a cost perspective (serverless will either cost you more at the high end of the scale…
https://www.helionenergy.com/ https://cfs.energy/ https://www.tokamakenergy.co.uk/ https://tae.com/
Insert below; jokes about its being 25 years away, and comments about how it will never work etc