Nvidia announced that it will launch NVIDIA AI Enterprise, an enterprise-level artificial intelligence integrated suite optimized, certified, and supported by Nvidia, exclusively for VMware vSphere 7 Update 2.
As the first industry cooperation, NVIDIA and VMware have developed an enterprise platform that supports artificial intelligence, virtualizing NVIDIA AI Enterprise's artificial intelligence workload on VMware vSphere. It provides companies with the software needed to develop a wide range of AI solutions, such as advanced diagnosis in the healthcare field, smart factories in the manufacturing field, and fraud detection in the financial services field. With this enterprise software suite, enterprise IT professionals who use vSphere for computing virtualization can use the same tools that manage large-scale data centers and hybrid cloud environments to support AI. Nvidia's software suite provides extended, multi-node, AI application performance on vSphere.
NVIDIA AI Enterprise provides compatibility for a large number of accelerated CUDA applications, AI frameworks, pre-trained models, and software development kits running in the hybrid cloud. Through optimization, workloads can be expanded across multiple nodes to support large-scale deep learning training models and achieve full GPU virtualization.
Nvidia also certified vSphere as the only computing virtualization software that can provide hypervisor support for live migration through Nvidia’s multi-instance GPU technology. This technology allows each A100 GPU to be divided into up to 7 instances at the hardware level to maximize Maximize the efficiency of workloads of all sizes.
In addition, the selected NVIDIA ConnectX adapters are now certified for use with VMware vSAN's RDMA (Remote Direct Memory Access), which offloads CPU communication tasks, thereby improving application performance and improving the return on infrastructure investment.
0 Comments