Research and analyst firms have several takes on the future of data centers, but the prevailing belief is that cloud computing will take a dominant role.
By Patrick McLaughlin
Market research, forecasts and analysis concerning data center technologies—including those that directly impact the deployment of structured cabling systems within these facilities—tell several stories about what the industry might expect from the data center business in the months and years ahead. Several research and analyst firms recently weighed in on one aspect or another of data center network administration.
Fiber in the lead
A 105-page data center cabling study carried out by BSRIA (www.bsria.co.uk) and published in November is based on data gathered from 335 data center managers across six vertical markets in the following countries: India, China, the United States, the United Kingdom, Germany and France. “In terms of copper and fiber structured cabling in data centers, these six countries make up close to 70 percent of the data center worldwide structured cabling value in 2010,” BSRIA said when announcing the study’s availability.
The study was conducted to look at the uptake of high speeds and to discover the media over which these speeds are being run, or plan to be run by the end of 2013.
“Overall in data centers surveyed in 2011 across the six countries, the copper-to-fiber ratio is 42 to 58,” BSRIA said, adding that the use of MPO-style connectivity and preterminated cabling is increasing. “There is some ambivalence toward the decision between using structured cabling or point-to-point links for future 10G copper links,” BSRIA noted.
The report is accompanied by an addendum that includes more detail on each country’s market.
40G connectivity boom?
Lisa Huff, an analyst who examines the high-speed networking, cabling and interconnect industries, recently reported on the market potential for data communications connectors and cable assemblies for 40G systems.
Huff, a telecom director with the research firm Bishop & Associates (www.bishopinc.com), recently wrote about the state of the market –from both technical and business perspectives—for 40/100G connectors, QSFP+ connectors and cables, and CSP and CFP connectors and cables for 100G Ethernet. In her reporting she noted, “As each new data rate has been released, the servers’ adoption of it lags by at least three years. And for 10G it has been more than eight years before the servers have been able to take advantage of the I/O speed. Server manufacturers were concerned that if the IEEE adopted 100G without a ‘speed bump’ in between, it may drastically slow its adoption. ... 40G was added, and we see early adoption of it, at least for high-end data center applications, only a year after the standard was released.”
At least some of the optimism for 40G may come from the dynamic that Huff observes in the 40G market. She says, “Traditionally Ethernet switches lag in the development of server technology when it comes to speed migration. This is because it is much easier to develop a chipset and optical module to fit on a traditional network interface card than it is to fit into a high-density switch port. We saw this with Gigabit and 10-Gigabit Ethernet, but it seems to be reversed for 40G. Several switch manufacturers have already released products with multiple 40G ports. The switches have 48 10G SFP+ ports for connections to servers and four 40G QSFP+ ports for uplinks to aggregation/distribution switches.”
10G adoption accelerates
A report from Dell’Oro Group (www.delloro.com) states that users readily consumed 10-Gbit Ethernet connections for their servers despite Intel having delayed the launch of its Romney server platform from late 2011 to early 2012. Dell’Oro said, “10-Gbit Ethernet controller and adapter port shipments grew 47 percent over the last year to about 1.3 million” in the third quarter of 2011.
Sameh Boujelbene, an analyst covering controllers and adapters for Dell’Oro, noted, “We are still in the early stage of 10-Gbit/sec server adoption and we expect that Intel’s Romney platform will fuel even more migration toward 10-Gbit in 2012. New entrants to the Ethernet server connectivity market are carving out specialty niches and are changing the supplier landscape regularly. We expect to see continuous changes throughout the next couple of years.”
In a separate report, Dell’Oro concluded that the Ethernet switch market is undergoing a technological transformation in the data center, as virtualization changes where and how applications are connected to end users. The firm stated, “Two major trends will forever alter the Ethernet switch market: 1) a significant technology change toward 10-GbE and fabrics within the data center, and 2) the emergence of powerful new customers.”
Alan Weckel, senior director with Dell’Oro, said, “Enterprises are consolidating data centers and are moving infrastructure toward the hosted environment of the cloud. These migrations toward the cloud are reducing the numbers of both data centers and IT end users that are making discrete purchasing decisions, as fewer but larger equipment purchases are needed. Over the next five years, vendors will expand and consolidate as the battle for supremacy in data center networking intensifies. The result is that there has never been a better time for new entrants or a better opportunity for existing vendors to gain share.”
That analysis dovetails with the findings of a document recently published by Cisco Systems (www.cisco.com). “The Cisco Global Cloud Index: Forecast and Methodology, 2010-2015” is chock full of data and projections from the networking giant. Cisco describes the document as “an ongoing effort to forecast the growth of global data center and cloud-based IP traffic.” The current version is 26 pages in length.
Among the projections within it is, “The year 2014 is expected to be a pivotal year, when workloads processed in cloud data centers (51 percent) will exceed those processed in traditional data centers (49 percent) for the first time.” The report adds that, “we expect cloud-processed workloads to dominate at 57 percent by 2015.”
The index defines a workload as “the amount of processing a server undertakes to execute an application and support a number of users interacting with the application.”
Another fact of note from the index is that most data center traffic stays within the data center from which it originated, rather than flowing from one data center to another or flowing from a data center to end users through the Internet or IP wide area network.
“In 2010, 77 percent of traffic remains within the data center, and this will decline only slightly to 76 percent by 2015,” the index says. It cites the following three factors as primary contributors to traffic remaining within data centers.
- Functional separation of application servers and storage, which requires all replication and backup traffic to traverse the data center
- Functional separation of database and application servers, such that traffic is generated whenever an application reads from or writes to a central database
- Parallel processing, which divides tasks into multiple smaller tasks and sends them to multiple servers, contributing to internal data center traffic
According to Cisco, virtualization offsets video, thereby maintaining the high level of internal data center traffic. The index reports specifically: “The ratio of traffic exiting the data center to traffic remaining within the data center might be expected to increase over time, because video files are bandwidth-heavy and do not require database or processing traffic commensurate with their file size. However, the ongoing virtualization of data centers offsets this trend. Virtualization of storage, for example, increases traffic within the data center because virtualized storage is no longer local to a rack or server.”
These factors promise to keep the data center marketplace busy and dynamic for years to come.
Patrick McLaughlin is our chief editor.
Past CIM Articles