Therefore, the PSU inside an AI server must supply more power than regular servers to drive these high-performance components, while also ensuring
Utilizing high-efficiency MOSFETs, sophisticated gate drivers and dsPIC DSCs equipped with high-performance and advanced peripherals enables
Conclusion Fulfilling the power demands of massive AI workloads, complying with the established CRPS form factor, and meeting 80 PLUS
To support GPUs, AI servers require 3~10 times higher power than traditional servers The GPU, which supports 48 V, has changed the output of PSU from 12 V to 48/54 V and has become the
An AI data center server power supply built using devices such as GaNSafe can achieve significantly better performance and support enhanced system safety and reliability versus a unit that utilizes
The growing demand for power in AI applications has created a pressing need for power conversion solutions that are both highly efficient and compact. To support the development of next-generation
The proliferation of AI has significantly reshaped data center infrastructure, pushing the limits of power systems to meet unprecedented
The new system architecture significantly improves energy-efficient power distribution in the data center and enables power conversion directly at the AI chip (Graphic Processing Unit, GPU) within the server.
High Efficiency, Compact DC/DC Regulators Optimize Power Delivery ited for AI server power architectures. Models such as the SiC461, SiC431, and SiC450 offer wide input voltage ranges, high
Discover how AI features like "Hey Siri" rely on powerful data centers. Learn about the technology behind smart factories and the importance
New architectures and AC-DC distribution configurations are increasing demand for data center rack and PSU power, necessitating more processing power. This article examines some
Learn how to new Keysight high-power ATE system power supplies deliver superior density paired with robust automation software not available from any other vendor – so engineers can validate complex
When operating hyperscale AI data centers, 120 kW of rack power is required. The efficiency to convert grid power to the voltage for the GPUs is about 88%, so it produces 15 kW of wasted heat that must
May 20, 2025. Infineon Technologies AG is revolutionizing the power architecture required for future AI data centers. In collaboration with NVIDIA, Infineon will
As server power demand is set to increase 2x-3x, designers now face an even tougher challenge to satisfy the demands for greater power delivery and significantly higher efficiency within the CRPS
Explore how innovations in power devices, gate drivers, and DSP-based controllers tackle AI servers'' high energy demands, optimizing efficiency in data centers.
Scope and purpose This document introduces a new, complete power supply unit (PSU) for AI data centers and servers.
Between 2024 and 2026, strategic momentum in AI high power server power supplies increasingly centered on enabling higher power density and better thermal/electrical performance for
The ever-increasing power demand driven by AI data centers is forcing an expedited evolution of power supply units (PSUs) designs, growing from 800 W to an astounding 12 kW, with projections heading
Hybrid TCM/CCM control strategy offers a comprehensive approach, combining the strengths of both modes to achieve higher efficiency, performance, and reliability in high-power AI server PSUs.
The AI High Power Server Power Supply Market was valued at USD 5.60 Billion in 2025 and is projected to reach USD 19.53 Billion by 2032, growing at a CAGR of 19.5%.
Powering the Next Generation of AI Hardware Partner with engineering teams to invent next gen solutions Nimble product development process allows customers
Contact us for competitive quotes on any of our power communication and smart grid products
Get a Quote