Cisco announces updates to the servers and software that comprise its Unified Computing System (UCS), catering to customers that need to run demanding applications.
The M5 servers are built on Intel’s new Xeon Scalable processors and capable doubling the memory capacity of previous systems. The UCS M5 servers offer up to double the memory capacity of previous systems up to 86 percent higher performance.
Specifically, the fifth-generation UCS servers include two new blade servers and three new rack servers.
We can bring customers these types of powerful new machines in a completely plug-and-play environment, said, Todd Brannon, director of product marketing at Cisco for Unified Computing.
The M5 generation servers, which leads the industry in GPU density up to two GPUs. Along the B480 M5 Blade Server for workloads ranging from memory-intensive enterprise applications distributed database virtualized workloads.
The C220 M5 Rack Server is a high-density, two-socket rack server. While, the C240 M5 Rack Server is a storage. I/O optimized enterprise-class rack server for big data analytics, software-defined storage and bare metal applications. Lastly, the C480 M5 Rack Server features an innovative modular architecture for flexible technology refreshes.
Google ‘PAIR’ improves AI and Machine learning
GPU support has tripled with up to six supported as having disk capacity, which now supports 32 drives. Additionally, Cisco extends its infrastructure automation and optimization capabilities with updated software. The updated software should help organizations that face workloads which are increasing at a far faster pace than IT budgets.
UCS Director 6.5 extends automation capabilities beyond infrastructure by automating native PowerShell functions, virtual machine mobility across vCenter data centers and support for VMware VMRC console.
The UCS platform allows the software to go through very deep levels of inspection, enabling customers to increase workload density at least 25 percent. Also, enables customers to plan ahead, accurately assessing how many M5 servers they would need to for optimal performance or what workloads should go to the cloud.