Supermicro is prioritizing the speed of AI factory deployment and data pipeline integration over raw GPU counts to help customers bring clusters online faster.
A former Supermicro contractor has pleaded guilty to charges involving the illegal diversion of Nvidia-powered AI servers to China in violation of US export controls.
The Supermicro Open Storage Summit focused on production-ready AI infrastructure, highlighting the role of storage architecture in optimizing GPU utilization and financial capital efficiency.
Supermicro is partnering with SpaceXAI to build a gigawatt-scale data center, leveraging Nvidia's latest hardware to meet ambitious computational targets by 2027.
Supermicro has started shipping NVIDIA Vera Rubin NVL72 rack systems, offering scalable data center solutions with liquid cooling and support for up to gigawatt-level power.
Supermicro has launched shipments of its Vera Rubin-based server racks, featuring 1,152 GPUs per NVL4 configuration and scaling to gigawatt-class AI clusters.
Broadcom and Supermicro are partnering to provide a unified AI infrastructure management solution for enterprises and cloud providers.
Enterprise AI implementation is moving beyond model selection, with companies now focusing on data accessibility, infrastructure optimization, and sovereign AI to secure proprietary assets.
Early benchmarks of the 5th Gen AMD EPYC server on Linux 7.3-rc1 show performance gains over the stable Linux 7.2 kernel.
