AI applications in data centers are quickly advancing, leading to a surge in the development of new hardware accelerators for Machine Learning (ML), Deep Learning (DL), and High-Performance Computing (HPC). These new devices have high power demands, thereby assigning data centers a critical role in global power consumption. Updated regulations and technological improvements, including on efficiency, will be crucial to moderate the surge in energy consumption from data centers. The power conversion chain is crucial for maximizing overall system efficiency. This paper introduces an innovative power conversion architecture that employs a multi-stage approach to ensure high efficiency and a rapid transient response, thereby enhancing Power Density and Power Delivery Network (PDN) performance.
Novel TLVR Operation In Multi-Stage Voltage Regulator Module With Current Multipliers
Zufferli, Kevin
Primo
;Rizzolatti, Roberto;Ursino, Mario;Mazzer, Simone;Saggini, Stefano
2025-01-01
Abstract
AI applications in data centers are quickly advancing, leading to a surge in the development of new hardware accelerators for Machine Learning (ML), Deep Learning (DL), and High-Performance Computing (HPC). These new devices have high power demands, thereby assigning data centers a critical role in global power consumption. Updated regulations and technological improvements, including on efficiency, will be crucial to moderate the surge in energy consumption from data centers. The power conversion chain is crucial for maximizing overall system efficiency. This paper introduces an innovative power conversion architecture that employs a multi-stage approach to ensure high efficiency and a rapid transient response, thereby enhancing Power Density and Power Delivery Network (PDN) performance.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.