University of Michigan Software Halves AI Training Energy Consumption
Researchers at the University of Michigan have developed Kareus and Perseus, two software tools capable of reducing AI model training energy by up to 50%. NVIDIA, a prominent AI computing company, is currently evaluating the integration of this energy-saving software into its open-source training platform, Megatron-LM. This innovation addresses growing concerns about the substantial power consumption of large-scale AI model training, offering a potential solution for greater efficiency.
Want more?
Open NewsSnap.ai for the full app experience, including audio, personalization, and more news tools.
Open NewsSnap.ai