Equinix Extends Global Data Center Footprint

With cloud computing leading to the deployment of distributed applications on a global basis, the location of data centers is often just as important as how much compute power it can provide. After all, the major bottleneck when it comes to deploying these applications is network latency.

To rise to that challenge Equinix (EQIX), a provider of hosting and network services, has already made $7 billion in IT infrastructure; including $227 million announced today that was used to build additional data center facilities in Secaucus, N.J., Singapore, Melbourne, London and Toronto.

Karl Strohmeyer, president of the Americas for Equinix, notes that different types of application workloads requires different approaches to how those applications are hosted in the cloud. Web applications tends to be very latency sensitive, so some portion of that workload needs to be hosted as close as possible to where that application is being consumed to minimize network latency. Other applications, such as batch-oriented big data analytics applications, can more often than not be hosted in a centralized data center.

Of course, other factors such as compliance can determine where workloads need to run. Many countries have enacted data sovereignty laws that require at least the data to resident within that country. That requirement, however, is in turn fueling interest in new forms of hybrid cloud computing where the data is stored locally, but all the compute processing is provided via a data center managed by a cloud service provider that could be anywhere at the other end of a virtual private network (VPN).

Click here for Talkin’ Cloud’s Top 100 CSP list

Strohmeyer said Equinix, which sells direct and via the channel, is most interested in workloads that have a high degree of sensitivity to network latency. In fact, the most recent fourth quarter Equinix sold more cross connects inside its data centers in its history, said Strohmeyer.

The primary issue that most IT organizations now face in the age of the cloud is that many older data centers were built in places that provide inexpensive access to real estate and power. But in the age of the cloud, IT organizations have rediscovered how critical networking is to the performance of cloud applications. As a result, more workloads are starting to be deployed in data center facilities that are not only one hop away from an Internet peering exchange; they also provide access to multiple carriers that IT organizations can play off of one another to get the best possible pricing.

Once those workloads move into the hosting facilities it’s also not uncommon to see IT teams moving into offices either in the facility or in offices near buy to be physically closer to where their IT infrastructure equipment resides, said Strohmeyer.

The end result is a fundamental shift in not only where IT infrastructure is located, but also via recent alliances with providers of IT monitoring software such as ScienceLogic along with Cisco for network management and Apigee for API management software how it needs to be managed.