128GB seems the sweet spot for local models. I can program and install most GitHub projects with opencode and QWEN 32b with mtp.

anyone whose addicted to token theoughput is losing the operational knowledge and offline capabilities.

if you arent moving to the AMD 395 or MACs then youre hitching aride on the expensive calory ride

▲

throw1234567891 2 hours ago | parent [-]

If you could buy a 256GB you’d be claiming that 256GB is a sweet spot. But I agree with you. Crack-tokens are not the future.

▲

cyanydeez 2 hours ago | parent [-]

no, the fact that MACs and x86 and soon ARM are all going to have 128GB models in every sector, yeah, sure.

But watching everyone flounder because claude goes down or forcing you on API costs.

I'm programming things that'd take me days with a PC that, without OpenAI's VRAM shenagans, would cost you $2k.

It's more than just 'this is what I could do' it's definitely about 'this is what anyone could do with a new PC purchase'.

▲

throw1234567891 2 hours ago | parent [-]

You must be unaware that System76 was already selling 192GB machines, mac studios used to be 512GB max. The only reason why we don’t have them anymore is that we are in RAM shortage.

▲

cyanydeez an hour ago | parent [-]

I'm aware you can have more. the term "SWEET SPOT" references a area that anyone/everyone can get to and isn't some magical expensive unicorn.

You're doing what the IT industry has been addicted to for decades: number goes up.

	▲	throw1234567891 7 minutes ago \| parent [-]
		> You're doing what the IT industry has been addicted to for decades: number goes up. No, I have a hands on experience with bigger models, and understand the advantages of using them.