Hacker News new | past | comments | ask | show | jobs | submit login

Repair-friendly form factor for a "Unified Memory" platform.

For $2000, you get 128 of system RAM, 96 of which is addressable as VRAM. Only ways of getting 96GB of VRAM in a desktop are to either:

1. Drop ~$5000 on a (very non-upgradeable) Mac Studio 2. Drop ~$20k on a dual RTX 6000 workstation

For running local LLMs, there's nothing on the market presently even remotely like this.




You can get a M3/M4 Max with 128 GB of RAM as well. The Studio will give you > 128 GB.

I have a max with 64GB RAM, which is good enough for 70b models with a 3 bit quant. Even if I had more RAM to run larger models, my GPU would be the bottleneck.


> You can get a M3/M4 Max with 128 GB of RAM as well.

To get an M4 Max so you can have 128GB, you need a macbook pro. The cheapest macbook pro with 128GB is $4700.

M4 Max does have the benefit of more memory controllers, so it has twice as much memory bandwidth as Ryzen AI Max. But that's a lot of money to pay for it.


To get an M4 Max so you can have 128GB, you need a macbook pro.

By the time the Framework ships the Mac Studio will have been updated to the M4 Max. Although 128GB will still probably be around $3k.


Yep. You can get an M3 Max refurbished, but will make some tradeoff with GPU performance.




Consider applying for YC's Summer 2025 batch! Applications are open till May 13

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: