A new analysis from the AI Infrastructure Insight series argues that faster GPUs alone cannot deliver AI performance, citing a UC Berkeley, ICSI, and LBNL paper showing that over the past 20 years, pe…
A study by researchers from Seoul National University, University of California, Berkeley, ICSI, LBNL, and University of Texas at Austin finds that Diffusion Language Models (DLMs) achieve higher arit…
UC Berkeley, ICSI, and LBNL researchers introduced ARBITRAGE, a step-level speculative decoding framework that uses a lightweight router to dynamically choose between draft and target model steps, red…
A developer created AI Meter, a local token usage tracker that estimates energy and water impact from AI coding-tool logs, after discovering their personal AI usage consumed nearly 1 MWh in the past t…