44 ms·
On my 24GB RAM M4 Pro MBP some models run very quickly through LM Studio to Zed, I was able to ask it to write some code. Course my fan starts spinning off like
by giancarlostoro 8mo ago
On my 24GB RAM M4 Pro MBP some models run very quickly through LM Studio to Zed, I was able to ask it to write some code. Course my fan starts spinning off like the worlds ending, but its still impressive what I can do 100% locally. I can't imagine on a more serious setup like the Mac Studio.
- efxhoy 8mo agoHow is the output quality of the smaller models?
- elsombrero 8mo agonot good enough for coding anything more than simple scripts. generally, the less parameters, the less knowledge they have.
- kraig911 8mo agowhat model were you using?
- giancarlostoro 8mo agoWrote about it here: https://news.ycombinator.com/item?id=47191915 https://news.ycombinator.com/item?id=47191915
- jbellis 8mo agoYour limitation after prefill is memory bandwidth. A maxed out Studio has less than a single 3090 (really).
- sroussey 8mo agoYeah, the 3090 has faster memory, but not by a lot. The 5090 is at 1,792GB/sec and potential M5 Ultra would be 1,230GB/sec and 512GB RAM. Maybe 1TB. Not 32.
- thejazzman 8mo agoYou’re suggesting that a difference of the entirety of the M5 Max’s bandwidth is an insignificant gap!
- veidr 8mo agoNo, that difference is the 5090, not the 3090.