AI technology faces a critical challenge: its significant energy consumption. Noel Hurley, former VP at Arm, aims to address this through Literal Labs by innovating beyond the traditional neural network approach. Literal Labs employs Tsetlin machines, leveraging propositional logic for more efficient data processing, which promises to be 10,000 times more energy-efficient with added hardware acceleration. This new method, developed from patents by cofounders Dr. Alex Yakovlev and Dr. Rishad Shafik, offers a scalable solution for industries such as mobility, robotics, and transportation. Hurley, drawing from his experience at Arm, focuses on partnerships and broad chip compatibility to enhance market accessibility. This strategy mirrors the successful approach at Arm, prioritising power efficiency and fast inferencing to meet the pressing needs of modern tech applications.
'AI isn't energy efficient right now. This ex-Arm VP has joined a new startup to try to change that.' can be read in full at Business Insider.
Authored by: Riddhi Kanetkar
Published: 16 Feb 2024 by Business Insider