FlexPod AI

2019: Submission of the solution

On February 7, 2019 NetApp, the company in the area data storage in hybrid cloud, reported that together with Cisco FlexPod AI — the convergent solution constructed on widely scalable highly productive a flash infrastructure in combination with servers Cisco UCS 480 ML M5 and graphic provided processors Nvidia. FlexPod AI is the solution for customers who appreciate technical support and the architecture of FlexPod which highly proved in the market. Besides, the solution will be relevant for persons interested to expand already available infrastructure FlexPod to support the last types of workloads, such as applications artificial intelligence, machine and deep learning.

The presented solution allows to manage data on the periphery, in a core and in a cloud and also to surely apply frameworks to AI. At the same time the high performance of FlexPod AI is reached at the expense of the optimized calculations of the graphic processor and widely scalable flash.

FlexPod is a stack of convergent infrastructure which supports end-to-end NVMe when using Cisco MDS and NetApp AFF A800. In an array of AFF A800 NVMe disks while the NVMe over FC technology is supported by the UCS servers, switches of MDS Fibre Channel and DWH of Netapp are set. It provides the low latency necessary for needs of Real-time of applications.

Main information on release:

• FlexPod AI using UCS 480 ML M5 with support of graphic processors;
• End-to-end NVMe with Cisco MDS and AFF A800 for low latency;
• Modern applications, including SAP HANA, SQL 2017, Oracle 12c, VDI with GPU;
• Components of the last generation: UCS M5, 100Gb Nexus and ONTAP 9.5 for the increased performance and efficiency.