Artificial intelligence (AI) is driving unprecedented demands in AI server power consumption, pushing the limits of data center infrastructure. In this video, Bill Wong, senior content director at Electronic Design, speaks with Maurizio Acosta, applications marketing manager at Infineon, about the challenges and solutions in powering high-performance GPU data centers.
As high-power AI servers continue to scale, the need for efficient power conversion becomes critical. AI workloads rely on GPUs that perform complex calculations in parallel, consuming significant energy and generating heat. A single server rack can exceed 120 kilowatts, requiring advanced cooling solutions. Companies are turning to innovative semiconductor technologies like Silicon Carbide (SiC) and Gallium Nitride (GaN) to improve efficiency and reduce wasted energy.
For more industry-specific information, watch the full video.
