News for hackers. Stuff that matters.
Re:The RAM shortage comes for us all by agoodusername63 on Thursday December 04, 2025 @05:56PM (#46154353)
isn't that what the mixture of experts trick that all the big players do is? Bunch of smaller, tightly focused models
Re:The RAM shortage comes for us all by irthomasthomas on Friday December 05, 2025 @11:57AM (#46163926)
Not exactly. MoE uses a router model to select a subset of layers per token. This makes them faster but still requires the same amount of RAM.