-rick-

Just spend 5 years of subs money on a mac to run the LLM on and hope that in 2 years time it's still good enough to run the current models Reason I asked is you said the new setup seemed to be performing better and everything I've seen suggests that local models are generally a little worse at chat type stuff and quite a bit worse for agentic workflows (still much better than the leading models from 9 months ago). I guess with the $20 gpt sub you can't access the better OpenAI models?

Which means? ChatGPT? Free, Plus, Pro? Codex? Claude?

What was your previous setup?

Oh.... You are still working that way? I thought that was considered very inefficient these days. Better to have it help you write a spec, including testing regime, code-quality metrics, design philosphy and then ask it to go build (and spawn agents/tools to help). Not that I'm doing any of this. Just watching people like Theo explain it: https://www.youtube.com/@t3dotgg/videos Had this in my 'to watch' list for a while:

DHCP? Either change it to a static IP in the settings (could cause trouble if you change internet routers) or go in your internet routers control panel and assign it a fixed IP there (DHCP will then always dish out the same address).

Does all the normal stuff work without the cloud? Full ability to change configs, etc? This situation is why I'm very dubious of SigEnergy. They seem to be market leading technology wise but they are very cloud dependant and I don't want to put money into a system that could go away/start costing more than expected/lose features.

Good job Macs are about the most efficient AI platform then

You looked at Gemma 4? Supposed to be able to get qwen like performance/capability but in a much smaller model. 96GB bit of a waste for it

I think they are just answering a different question and expect someone who is working with electrics to know what they are looking for. They are just talked about current carrying capacity of the cable. Up to you to work out how much current your devices will need. Yep so some derating required. 13A socket is not rated for 13A continuously. But kettles are only on for a few minutes before they boil. I would guess a lot less than that. My oven seems to warm up pretty fast and it's not a fixed install (got a 13A plug on the back) so isn't the most powerful. One of these threads triggered me to look up the ratings of fixed install ovens and for single ovens they are all saying they need 16A breakers. So if they are installed on a dedicated circuit with a 16A breaker there is no diversity applied. (Electrician will still do calcs for sizing upstream components). Sounds about right (caveated that insulation/distance derating might drive requirements higher), though my own thought is to always use 4mm as that allows the potential to switch from a single to double in future without rerunning wires. (4mm might not be enough for a double if big derating factors are required). I missed that your ovens are fixed installation in the other thread. If they don't come with a 13A plug (and are therefore fused) they need a dedicated circuit (as the dedicated circuit provides the fuse).

Do you not have separate controls for each ring on a hob? If you only have a single point of utilisation, all the rings turn on when you turn one on? Modern induction hobs usually have ratings for the individual rings and even break out groups of rings to different wiring terminals (to allow you to split the load over multiple phases).

We are such delicate flowers these days! Don't get me wrong, if I get to build I'm getting the cisterns with built in extract to remove the smells at source but to have this as a notable selling point for a £5-10k installed item makes me laugh.

Forgot to say, this does seem like it would be a bad thing if it applied to balcony solar type setups. If people have to pay engineering fees to plug in a device to their wall outlet whats the point in the changes on balcony solar? But the wording suggests it applies to only some situations which I would guess 800W of balcony solar is not going to be one.

Good question Joth. Either way, I don't find this idea bad. Seems reasonable to me. If you are spending £5000+ on solar (and given currently pricing are going to get payback for that very quickly) why should the rest of us sponsor the connection? If solar was a money loser for people and just done for environmental reasons then sponsorship makes some sense but given solar is so profitable that HMRC are going around reminding people profits from home solar generation is reportable income then further subsidy to higher income people seems wrong. Not against providing incentives for lower income people to get solar. That feels like a different situation.

Depends on your device. Expensive ones can do some form of variable power, though I would call PWM a smart form. The dumb control is bang-bang control (https://en.wikipedia.org/wiki/Bang–bang_control) But again this is irrelevant to the discussion. Even the bang bang control leads to the average current draw for the hob being much lower than the max draw. Again, if there are short periods where everything aligns and every hob is on simultaneously for some seconds that's accounted for.

That's the point. Even if every ring is 'on' most of them won't be drawing much power. Most will be drawing a fraction of their rating to maintain the temperature. The only time a ring will draw maximum power for more than a handful of seconds is after you first switched it on and say you have a pan of water you want to bring to the boil. Then that ring will be on for like 2-3mins continuously. It is exceedingly unlikely that your wife (or anyone) goes into the kitchen, gets out six pans and puts them all on the hob at the same time and sets them all to max. Even in that case, the rating on a circuit is designed for cable protection, ie, a 32A circuit can sustain a 32A draw continuously over many hours on the hottest day of the year without melting/starting a fire/etc. The cable (and the breaker) can sustain much higher currents for short periods of time and that's what these diversity calculations take into account. They allow for a short period of very high usage (much higher than rating) so long as the average over time is below the rating by enough.