Supermicro Accelerates AI and Deep Learning from the Data Center to the Edge with New NVIDIA NGC-Ready Servers
AI is helping to solve some of the world’s most complex problems. Solving these enormous challenges require the computation of large amounts of data and highly optimized AI models running at scale. NVIDIA GPU Cloud (NGC) is the GPU-accelerated software hub for optimized AI and HPC. Supermicro’s NGC-Ready systems make it easy and efficient to run large workloads with a complete end-to-end NVIDIA Tensor Core GPU-accelerated hardware and software stack.
Time to market is key to success for today’s AI development. By using Supermicro designed NVIDIA GPU systems with all the latest AI stack installed and supported, data scientists and AI developers can start testing and training their AI models for product development and research.
The SuperStorage 6019P-ACR12L+ is a 1U server designed for organizations that need a solution for high-density object storage, scale-out storage, Ceph/Hadoop, and Big Data Analytics. This server is highlighted by Supermicro’s X11-DDW-NT motherboard family, which features support for dual-socket 2nd generation Intel Xeon Scalable processors (Cascade Lake), up to 3TB of ECC DDR4-2933MHz RAM, and Intel Optane DCPMM.
The promise of fifth generation (5G) wireless networks has been dangled in front of businesses for years. Large IT vendors and mobile carriers have been promising next-generation solutions that are poised to enable dozens of new use cases that will transform the enterprise. Finally, at the end of 2019, some of the power of 5G began to unlock as carriers began offering service in select markets.
The trials will run in Brazil, Germany, Spain and the UK this year and involve Altiostar, Gigatera Communications, Intel, Supermicro and Xilinx.
This white paper is intended to help an organization deploy an on-premises SUSE CaaS Platform cluster on Supermicro’s BigTwin™, Ultra and SuperStorage systems with the 2nd Generation Intel® Xeon® Scalable processor to support the latest, Kubernetes compatible workloads.