Chart bonanza by @ArtificialAnalysis! When @grmcameron and @_micah_h state that they’re pulling results charts, they’re not kidding! Llama 3.1 405B @ 132 T/S Llama 70B @ up to 570 T/S 10X faster inference than GPUs Start developing http://
cloud.sambanova.ai

→ View original post on X — @sambanovaai,