Cisco unveils unified computing system

MANILA, Philippines - From expensive to economical, complex to simple, rigid to flexible, isolated to shared, power hungry to energy-efficient, the data center is changing and Cisco is striving to lead the way in its transformation.

In an 11-city press launch through the Telepresence, the Cisco videoconferencing technology that allows “face-to-face” meetings over the network, the company announced a major transition in its data center architecture, which it calls Data Center 3.0, that seeks to deliver on the promises of “virtualization through unified computing.”

The goal, according to Pete Nichols, business development manager for data center, and Andre Smit, managing director for data center sales, is to come up with a self-managed and holistic approach to computing by simplifying the data center.

“The data center is transforming into an agile and efficient networked environment. With this evolution, the architecture of the data center is changing,” explained Smit in a presentation.

To be able to unleash the potential of this data center, there is a need for a unified computing strategy, or a cohesive system that unites computing, network storage access and virtualization. “The benefits include reduced total cost of ownership, increased business agility and improved energy efficiency,” he said.

Smit disclosed that Cisco started simplifying its data center infrastructure three years ago and the technology timeline for its full-scale introduction is coming to a close with its deployment in the second quarter of this year.

Among others, in 2008 and 2009, Cisco has introduced the Nexus 7000 (Data center Ethernet), the Nexus 500 (unified fabric), the Nexus 100v (VN-link), and Nexus 2000 (Fabric Extender).  

“A server is a server is a server...” said Nichols. “But when you add virtualization, it takes away a lot of complexity.”

The new Cisco data center is a cohesive solution with embedded management tools and unified fabrics, and does away with unnecessary switches, adapters and management modules, which allow the system to be used as a single device.

When compared to an existing legacy system for a large enterprise, for example, that has around 320 servers, an estimated $21 million in capital expenditures and $800 annually on power and cooling for three years, around 3,520 cables and 31 racks, the new data center architecture that Cisco is launching under a unified computing system still has 320 servers but will take only around $12 million in capex (or 43 percent savings), and $650 million on power and cooling (or 19 percent savings). Moreover, it has only 480 cables and 12 racks.

The benefits are clear, according to the two company executives. On average, it reduces capex to around 20 percent, and operating expenses to 30 percent.

While Cisco bared collaboration and partnerships with various IT companies such as BMC Software, EMC, Emulex, Intel, Microsoft, NetApp, Novell, Oracle, Qlogic, Red Hat, VMware and Accenture, it announced that it would also be enter the blade server market, along with its entry into the unified computing system model.

Smit explained that the value of this is in the integration process where Cisco’s own blade servers are standardized and embedded into the system.

Nichols admitted that the data center is a conservative environment and it will take months for adoption. For this, Cisco calls UCS as the “data center of tomorrow.”  

“We fully intend to lead in this transformation in the data center,” said John Chambers in a video message. “This new unified computing system brings together the concepts of compute, network, virtualization and storage in a way that we think isn’t just a product announcement but as the future which others will build.”