Give it a shot, 3.1 live one in AI studio/API and max out reasoning - not the one in Gemini app it’s an older model.
Another option is to use pipecat with their VAD and separate STT and TTS and any (fast) LLM of your choice - but it’s more plumbing and not a true speech to speech model
Yes the voice part of OpenAI realtime/voice mode is great but it’s pretty dumb compared to newer models and often gets stuck repeating itself.
Google’s Gemini flash live 3.1 is better, especially used via the API - it can do tool calling (including to other, even smarter LLMs if you set it up yourself), you can set the reasoning level (even high is still close enough to realtime) and it can ground answers in google search. I love bidirectional voice and right now it’s probably the best option. You can try it in AI studio
Easy install, discord/whatsapp/tg out of the box. And some agent orchestration out of the box where the main LLM can farm out tasks to different models/agents - yes Claude code has some of this too but I think this has more
put the APP3 through a washing machine recently by accident and they had the high pitched feedback if NC was used until they were fully dry after a couple of days. Have also put APP2 through washing machine before but never had this
Give them enough time and they will. EUV will hit limits anyway in a decade.
For china it's DUV+packaging for now, NIL/DSA mid-term, and MoS₂/2D chips long term. But wafer scale, defect free 2D logic is 20–30 yrs out, so no EUV shortcut anytime soon
Good idea. Are the articles teaching how to use your software to solve an issue people have, or do you just mention/link to the the software in the end of the article?
That is a good catch, double speed FP16, so FP32 perf should be half that on the X1. While useful for some things (like neural network inference and I guess some shaders) it's not directly comparable - NVidia style marketing
I find the Jetsons really overpriced... I can pick up a TX1 Nvidia shield TV for £150 with controllers (and even install Ubuntu on it), but the Jetson dev kit with the same SoC is £500? It's like they don't want people to start using them
Nvidia X1 already is 1 tflops GPU and the X2 (which is likely to be in the nintendo) is going to be around 1.5 tflops. Those are mobile SoC's with CUDA cores, fit OK in tablets (Pixel C has one), X1 peak power consumption is about 10 watts.
Xbox one GPU is 1.3 tflops and PS4 is 1.8 tflops. So in the same ballpark
If you don't believe in what you're doing then I think it could have been a mistake to take the money and carry on.
If the team and investors believed in the product, perhaps you could have asked if some of the current team were willing to take it on, and make it work. You could have retained a bit of equity for the year and the hard work you put in so far without having to commit any longer yourself.
NPUs - definitely a good use case for at least part of it, there are ports of whisper that use coreML/ANE with less power and 3x speed of CPU only