Workflow
800VDC(800伏直流)电源架构
icon
Search documents
英诺赛科:与英伟达达成合作,联合推动800 VDC(800伏直流)电源架构在AI数据中心的规模化落地
Zhi Tong Cai Jing· 2025-08-01 09:49
Core Viewpoint - The company InnoPhase (02577) has announced a collaboration with NVIDIA to promote the large-scale implementation of the 800VDC power architecture in AI data centers, which is designed for efficient power supply in future megawatt-level computing infrastructures [1] Group 1: Collaboration and Technology - The partnership aims to leverage NVIDIA's new generation 800VDC power system, which offers significant advantages in system efficiency, thermal loss, and reliability compared to traditional 54V power systems [1] - The 800VDC architecture is expected to support a 100 to 1000 times increase in AI computing power [1] Group 2: Product Features - The company's third-generation GaN devices provide excellent high frequency, high efficiency, and high power density characteristics, offering a complete GaN power solution from 800V input to GPU terminals, covering 15V to 1200V [1] Group 3: Future Outlook - The integration of GaN technology with NVIDIA's 800VDC power architecture is anticipated to enable a leap from kilowatt-level to megawatt-level AI data centers in the coming years, ushering in a more efficient, reliable, and environmentally friendly era of AI computing [1]
英诺赛科:与英伟达达成合作,联合推动800VDC电源架构在AI数据中心落地
news flash· 2025-08-01 09:46
Core Insights - The company InnoSec has announced a partnership with NVIDIA to promote the large-scale implementation of the 800VDC power architecture in AI data centers [1] Company Summary - InnoSec has recently entered into a collaboration with NVIDIA, a global leader in AI technology, to advance the deployment of the 800VDC (800-volt direct current) power architecture [1] - This new power system is specifically designed by NVIDIA for future efficient power supply in megawatt-level computing infrastructure [1] Industry Summary - The 800VDC power architecture offers significant advantages over traditional 54V power systems in terms of system efficiency, thermal loss, and reliability [1] - The architecture is capable of supporting an increase in AI computing power by 100 to 1000 times [1]