The UCS product family just got larger, but before I introduce you to an exciting new product, let me tell you a little about the context for the design and the shortcomings in the industry that we set out to fix.
The computing landscape is evolving quickly as modern applications become more diverse in their needs. At Cisco, we’re always working to stay ahead of these requirements. There is an increasing use of scale-out architectures driven by the boom in data analytics, the continued growth of cloud scale computing – both on-prem and hosted, and new demands for computing at the edge and in private clouds. Whether it’s a service provider looking for maximum density in a cloud platform, a manufacturer modeling new designs, a retailer analyzing consumer trends, or a data scientist modeling a financial market, the ability to extract economic benefit from data and IT infrastructure is simply elemental to success.
A unique, high-density server architecture is being used for some of these use cases to power compute-intensive workloads. This is, generally referred to as a “multinode” system. These high-density servers are essential to efficient operations but have traditionally been operated in clusters separate and apart from general purpose computing. Until now, this division between high-density computing clusters and the rest of the data center has resulted in silos of technology, policy and management for customers. These silos are the bane of IT efficiency and security.
Our belief is that computing of any form should be managed as part of a larger, cohesive system. That’s why we say UCS isn’t just a server, it’s a system. This idea of providing different computing architectures (rack, blade, storage optimized, converged and hyperconverged) that are optimized to the unique needs of different workloads, yet managed as a singular system, is still unique to UCS and core to our design philosophy.
Today, I’m pleased to announce that we are adding a new form factor option to our portfolio, one designed 
to meet the needs of performance-intensive, scale out workloads: the Cisco UCS C4200 Series Rack Server Chassis and Cisco UCS C125 M5 Rack Server Node.
Maximize performance, not complexity
Much like our S-Series storage servers, which are modular and optimized for storage density, the C4200 chassis takes a modular approach, but optimizes for compute density. The C4200 chassis houses four server nodes. To start, we are introducing the Cisco UCS C125 M5 Rack Server Node, based on the AMD EPYC™ processor family. The C4200’s modularity, combined with the fabric-centric nature of UCS and advanced management, is unique in the industry and delivers several key advantages:
- Up to 128% higher processor core density compared to existing two-socket UCS M5 rack servers
- 33% more memory bandwidth compared to existing UCS M5 rack servers
- No new fabric and management silos for high-density computing vs general purpose computing
- Systems management from the cloud, with Intersight
UCS: One system for any workload in any location
As computing demand shifts from large, traditional data centers to include smaller, more distributed environments at the edge, the ability to mix form factors seamlessly in “micro data centers,” and to manage and automate operations from the cloud becomes vitally important. With a fabric-centric architecture, centralized, cloud-based management, and the ability, now, to add dense form factors and new capabilities to UCS, we can better address new technologies as they emerge and respond to emerging customers needs. All this without creating new islands of infrastructure or requiring forklift upgrades.
This long-term persistence and consistency of management and operations is the hallmark of UCS. These new C-Series multinode servers extend our UCS portfolio to offer greater efficiency for scale-out computing clusters, yet are seamlessly deployed and managed right alongside B-Series Blade Servers, other C-Series Rack Servers, S-Series Dense Storage Servers, and HyperFlex Systems.
With Cisco Intersight, the C4200 chassis and the C125 server nodes it houses can be managed uniformly with all UCS and HyperFlex systems from a single console connected to a SaaS platform. This, making it easier to deploy, scale and manage infrastructure at the core or across remote, multi-site environments. Further, Cisco Intersight simplifies and keeps IT infrastructure management evergreen with the frequent introduction of new features that only a cloud-based SaaS model can offer.
A New Partnership
Followers of the server industry and Cisco UCS will note that this is the first instance of an AMD-based server in our lineup. We chose AMD EPYC server processors for the first node of the UCS 4200 platform due to the architectural synergy with the workloads our customers want to power with this type of form factor. I’d like to thank AMD for their partnership on this product and welcome them to the Cisco partner ecosystem.
Learn More
The UCS C4200 and C125 will be available early in calendar Q3 of this year. We’re very excited to add this new option to the line-up and I’d like to give a big shout out to the customers who have provided us with fantastic feedback and encouragement through our early access program. Please follow us on @CiscoUCS and @CiscoDC on Twitter for additional news and to learn more on the UCS C4200 Series Rack Server chassis and UCS C125 M5 Rack Server Node read our deep dive blog.
Hi KD,
The key question that people (inside and outside Cisco) will have is :
Why is Cisco coming back into the multi-node server market after having abandoned it in a haste 2 yrs ago? What was the trigger to exit earlier and what has changed now?
Above content is purely pointing to the technical specifications or key positives in going with AMD based servers…
Adding to it, even the cost factor would have been considered as one of the parameter?