Zach Williams

product designer | full stack developer | hardware developer | songwriter/producer | building öra (Thoughtfully designed products for musicians)

[Zach Williams @0xZach]

@mhpaler Never stop never stopping.

[Zach Williams @0xZach]

literally my life since january 25 (first code push)

[Kr$na @krishdotdev]

The new development cycle in one image.

[Zach Williams @0xZach]

gpt 5.6 sol is getting real problems solved in my wireless audio software defined radio that fable couldn’t figure out and at a much faster pace @OpenAI

[Zach Williams @0xZach]

@signulll hard to do at a large scale/large area. you could do things like noise cancelling at your seat in an airplane though. a cone of silence

[Zach Williams @0xZach]

@signulll i did a science fair project with this idea in high school 20 years ago with parametric speakers. open-air noise cancellation

[Zach Williams @0xZach]

@dagorenouf pivoting into hardware rn

[Zach Williams @0xZach]

@p1nosaur solder

[Zach Williams @0xZach]

@BLCNYY same

[Zach Williams @0xZach]

fable is great. the problem is im out of usage and its only half way done with my task

[Zach Williams @0xZach]

@ArmSoM_Global Will the AIM7 be available in configurations with 32GB of RAM? I see on your website there is only one option which seems to be 8GB RAM and 128GB storage

[Zach Williams @0xZach]

@astrodotbuild Plug this into @flueframework for RAG

[Zach Williams @0xZach]

@ArmSoM_Global Understood. I see the wiki only mentioned Gemma 2 and 3, but if the RKLLM software can take the GGUF files and run them, that’s great.

[Zach Williams @0xZach]

@ArmSoM_Global Oh wow that’s better than I thought. Was looking at running Gemma 4 E2B, E4B and 26B A4B. I’m currently running E2B on my Jetson Nano. Would the MTP models would run on it?

[Zach Williams @0xZach]

@ArmSoM_Global Thanks for the explainer! This is really helpful to know. So is the RJ3588 better for non-LLM AI related tasks? Or if it does work with running LLMs, are there any tokens/s figures on different models?

[Zach Williams @0xZach]

@googlegemma Is anyone doing optimizations like this with a Jetson Orin Nano? Would love to see it

[Zach Williams @0xZach]

@jichiep Does this work in realtime? If so what’s the latency?

[Zach Williams @0xZach]

@ApolloCampari @Sajeel_Purewal Got it. So a better image generator. Will be cool if someone can create shippable hardware tools.

[Zach Williams @0xZach]

@SipeedIO This is what I’ve been waiting for. Gonna be nuts to see what happens in the next couple of years with embedded intelligence