News for hackers. Stuff that matters.

A comment on Qwen3-Coder: Agentic coding in the world. Parent comment

Re:Qwen3-Coder: Agentic coding in the world by Jayakumark on Tuesday July 22, 2025 @11:31PM (#44655489)
What will be the approx token/s prompt processing and generation speed with this setup on RTX 4090?
[ Parent | on HN ]
Re:Qwen3-Coder: Agentic coding in the world by danielhanchen on Wednesday July 23, 2025 @12:53AM (#44655816)
I also just made IQ1_M which needs 160GB! If you have 160-24 = 136 ish of RAM as well, then you should get 3 tokens to 5 ish per second.

If you don't have enough RAM, then < 1 token / s

[ Parent | on HN ]