Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

The cheapest option is two 3060 12G cards. You'll be able to fit the Q4 of the 27B or 35B with an okay context window.

If you want to spend twice as much for more speed, get a 3090/4090/5090.

If you want long context, get two of them.

If you have enough spare cash to buy a car, get an RTX Ada with 96G VRAM.



Rtx 6000 pro Blackwell, not ada, for 96GB.


Ah thanks.

The names are so good and not repetitious.

No not the RTX 6000. No not the A6000...


Thanks this is a great summary of the tradeoffs!




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

Search: