First, find VCC and GND. Verify peak voltages with an oscilloscope on the signal wires. Connect all wires, do some stuff in the app while you record.

Most software has analysis functions, if not, check the usual protocols.

If you have more then 5V on the wire, it may be Modbus or other industrial connection standards.

If found the correct protocol, baudrate and so on, I use for example esp32s to sniff in the wire and do a more useful long time capture

What kind of socket? Pass through or closed? I ma have schematics for this

Often, those are only serial/wifi bridges. Worth al look. Logic analysers you can get for less then 20 bucks as a cheap USB knockoff, but they work. If it is just serial, congratulations, that is a straight forward implementation. And in Europe, this does not break warranty btw

Na, that fine. I have seen cloud enabled power tools … Yes …

Pass. Uses pgp for encrypting your password and git for storing them.

Has mobile apps, browser plugins, desktop frontend and a bunch of cool plugins like rotation helpers

I often replace those WiFi models with esp32s. ESPhome on top.

Currently working in a full integration for our heat pumps. I freed ceiling lamps, space heaters, and more from their chinese cloud backdoor.

If I can get a device that respects my privacy I will buy that version. If not, I am a Journeyman of Electronics (yes, rly) I have my beloved oscilloscopes, logic analysers and 50 EPSs in stock.

Heat pumps are just ACs that heat. We can switch ours between heating in cooling

Me, having air air heat pumps for heating. Lucky coincidence they are not only very efficient in heating, they saved my life with 40°C out there

That’s long term planning. Respect. Visiting prides over 20 years ago, being an ally for decades, everything for 2028! This man is a genius!

Edit:

When you build your port in civi just to get the boost

Wait, just wait. Some day AI will save the climate. One day

No shit Sherlock

To avoid context switching on the GPU. OpenWebUi for example uses it for memory and title generation.

Those are not performance critical and background tasks, so instead of slowing down qwen, we just outsource this stuff to the NPU.

Edit: see here for more details

That’s what I meant with no clean. Your explanation is way better to understand for people not familiar with this feature. thanks!

ThinkPad and Dell have a bunch of Linux compatible notebooks.

If you are in a European country not being locked into apples ecosystem would be a major argument for me.

AMD Strix is an APU, optimized for AI. It is the cheapest option I am aware of to run bigger models at home. 2k for 56GB VRAM, and less den 300W total power Budget.

One could run smaller models. But for the context sizes required for research work, that is nearly impossible.

Also, external services, like openrouter, can be used to use models hosted in the cloud.

But for self hosted, you need something that can run models with at least 15GB of VRAM + Context. For comparison. Our highly quantized model uses 20GB of vram. For our 4 slots we need another 20GB on top of it (around 5GB for 254k tokens), making it 40GB.

Not imatrix or advanced quants, but yes.

But there are more then enough stock models for this task I would say. For specialized use cases though costum quant can be very very powerful

Besides marketing slides and claims of people profiting from hosted models, you do not need a 1Trillion parameter model

Openrouter is also nice for this. You can use real cheap models for embedding and the bigger ones for the actual research.

“Hey Claude, research for me current research to Nuclear Fusion. What are the biggest hurdles what are the next steps, and how promising is private research” enabling the research feature will give you a report, Fact checked (not clean but ok ish), and all the sources for it.

Claude will spin up a bunch of workers and search the web, following leads, and so on.

One of the few actual useful features of AI IMHO