Countless resources have been spent over several decades on solutions that drive improvements within the datacenter. While these investments have given us a considerable number of new innovations, they have tended to do so within silos of datacenter infrastructure and have tended to provide greater benefits to capital costs than operational costs. Server virtualization, Moore’s Law, or multi-core CPUs (just three examples) have undeniably driven considerable compute density within the datacenter. Meanwhile, storage efficiency technologies like deduplication, compression, or storage tiering have helped drive comparable levels of density within the storage market. While these and many other innovations have allowed datacenter teams to do more with each successive infrastructure refresh, they have also left IT professionals with technologies that must be managed and refreshed independently by silos of experts. This has brought us to a point of highly inefficient datacenter operations. Indeed, too many datacenter teams find themselves stuck on a refresh treadmill, where too many IT staff are spending too much of their time going from one refresh project to the next. Further still, these silos of datacenter resources tend to tie up valuable resources dedicated to datacenter floor space, power and cooling.
The result is that operational expenses have increased far more rapidly than capital expenses over the past twenty years. Looking at server costs as a proxy, IDC data tells us that every 1995 dollar spent on a server resulted in about fifty cents of spending on managing, powering and cooling that server. This CAPEX:OPEX ratio flipped over the following ten years when, in 2005, every $1 spent on a server resulted in $1.5 spending on managing, powering and cooling that server. The most recent data tells us that this CAPEX:OPEX spending gap continued increasing for another ten years when, in 2015 a dollar spent on a physical server resulted in whopping $3.91 of spending on managing, powering and cooling that server.
Interestingly, IT departments seemed willing or at least able to live with this ever-increasing gap between capital expenses and operational expenses until only recently. Today, organizations of all sizes are undergoing a once-in-a-generation change. They are looking to four critical technology pillars (social, mobile, analytics and cloud) to drive dramatic, company-wide digital transformations. These transformations are brought about by a need to leverage technology and data to establish deeper relationships with their customers, to better understand trends within their business, and to generate new sources of revenue. Simply stated, companies undergoing such enormous transformation can no longer tolerate an IT department that overspends on managing, powering and cooling datacenter infrastructure. This is especially when that money doesn’t provide much agility.
It is now clear that many companies are well aware of these trends and are increasingly investing their resources on infrastructure that breaks down silos of datacenter technology and experts with the goal of freeing up resources to support the critical transformation efforts mentioned above. Among the objectives: reduce the number of staff hours needed to manage the core pillars of datacenter infrastructure — namely compute, networking and storage. A complementary objective is to reduce the need for separate compute, networking and storage budgets. The outcome of these changes will be a datacenter that is far more agile and ultimately more operationally efficient.
One of the underlying technologies supporting this shift is hyperconverged infrastructure. Such systems leverage software-defined storage to provide enterprise storage services through the same x86 server resources used to run hypervisors and applications. These systems eliminate shared, networked storage systems by virtualizing the storage, and integrating storage and compute functions into a single server node (or a cluster of nodes).
As a provider of hyperconverged infrastructure, SimpliVity recently worked with IDC to survey its customers and see just how such systems were impacting datacenter capital and operational costs. In terms of specific budget line items, more than 60% of the 135 respondents said their organizations reduced hardware capital expenses, and costs for power and cooling, and 50% realized savings related to datacenter floor space and facilities. In addition, IT budget priorities shifted after deployment of the SimpliVity system: nearly 60% of IT budgets prior to deployment were spent on maintenance tasks, with the remainder spent on new technology. Those ratios shifted to emphasize new technology purchases after deployment of SimpliVity hyperconverged infrastructure.
It is clear that the times have changed. IT teams must be more agile and must reduce the amount of operational resources needed to run a datacenter and support their business’s digital transformation. IDC believes that hyperconverged technology can help in just this way. By eliminating the need for separate silos for compute, networks, and storage, and the dedicated staff that goes with those silos, hyperconverged infrastructure can improve efficiency and reduce costs in the datacenter, and help create more operationally efficient IT staff.