News for hackers. Stuff that matters.

A comment on Kimi K3 (2.8T) at 1 token/s on a MacBook Pro, streamed from four SSDs. Parent comment

Re:Kimi K3 (2.8T) at 1 token/s on a MacBook Pro, streamed fr... by andy_ppp on Tuesday September 08, 2026 @09:33PM (#49619644)
Except Taalas is much faster than GPUs, orders of magnitude so. They aren’t going to get 100x faster at inference any time soon!
[ Parent | on HN ]
Re:Kimi K3 (2.8T) at 1 token/s on a MacBook Pro, streamed fr... by SPascareli13 on Wednesday September 09, 2026 @06:21PM (#49635361)
There's a new SOTA model every few months, are you supposed to buy a new chip every new release?
[ Parent | on HN ]