The market demands on the data center industry seem to push multi-tenant data centers (MTDCs) and enterprise data centers toward higher density computing, efficiency, and increased speed to market. There are many solutions to this trend and deciding on the best one to adopt can be difficult to navigate.
Last year, Mission Critical Magazine featured an article that showed readers how to craft a total cost of ownership (TCO) model and analysis. This year, we will use the guide to demonstrate how multi-tenant or enterprise data centers can craft a TCO model and analysis on modular data center infrastructure, specifically using data modules versus a traditional raised floor design. By the end of this article, you will have a blanket understanding of modular data center infrastructure and its possible benefits, and you will be able to build a TCO model and analysis of data modules in relation to your own MTDC’s objectives to make an informed decision for your business.
FIRST STEP: BEFORE YOU BEGIN
Gather all costs associated with the adoption of modular data center infrastructure with data modules incorporated into the design, inclusive of both capital expenditures (CAPEX) and operational expenses (OPEX). CAPEX is composed of the hard cost of construction and equipment, as well as soft costs such as engineering, design, equipment lifecycle, and replacement costs, etc. OPEX is composed of costs like power consumption and other utilities, preventative and predictive maintenance connectivity, and staffing. After accumulating all the necessary data, you can begin building your model.
SECOND STEP: THE OVERVIEW
Introduce your TCO model and analysis of the proposed data modules used in the data center design vs. a traditional raised floor design. Detail any relevant background information, describing the company’s original build and intent, as well as any other pertinent history. Be sure to detail any statuses or numbers at the heart of the problem, especially any that will be relevant to the model’s comparison later on. This likely includes the data center’s current power usage effectiveness (PUE), which is a measure of how efficiently a data center uses its power. The goal of the TCO model and analysis is to steer your data center towards the best possible solution for your needs at the lowest possible costs.
THIRD STEP: THE GLOSSARY OF TERMS AND ACRONYMS
Define each of the technical terms and acronyms that will be used in the model. It is also helpful to give the data center raised floor design as it is now a term, like “Data Center A” or “DCA,” and the proposed data module design another term, such as “Data Center B” or “DCB.” This keeps the two delineated, and helps the writer quickly reference one or the other. Display how all of these terms and acronyms will be written going forward in the text for your reader’s reference.
FOURTH STEP: BACKGROUND
Describe previous solution attempts chronologically and why they were rejected. This section will vary for every data center, but the most crucial take-away from this history is the data center’s limitations and the resulting requirements for an effective solution. Lay out the steps that brought your company to consider data modules. Summarize the new design and estimate its impact.
FIFTH STEP: THE SCOPE OF THE MODEL
Explain that the TCO model’s purpose is to address data modules defined as the efficient housing for the customer’s IT hardware, network, and IT support equipment; not power/infrastructure modules nor containers, which are often confused for data modules. A data module is self-contained with efficient cooling, fire suppression, security, monitoring, and operating systems. Well-built data modules adapt to the customer’s supplied IT hardware, network, IT support equipment, and different racks. Mention any other designs your data center has researched and their projected yields. Develop and list selection criteria, for example: reduction of capital investment for construction, reduced PUE and OPEX, an increase in sellable units (e.g., more kW for customers in a colocation model), and more efficient use of the available square footage in the data center’s IT equipment white space. Rank the designs in accordance with the company’s long-term goals to showcase the best options. Summarize the model’s comparison of raised floor design versus modular design.
SIXTH STEP: THE DEMONSTRATION OF NUMERICAL FINDINGS
Show the CAPEX and OPEX totals for each solution in a spreadsheet. The numbers will support your conclusions. Ensure that you compare each design against a common amount of capacity in megawatts or kW for an “apples to apples” comparison.
SEVENTH STEP: THE PROS AND CONS ANALYSIS
For each option, make tables from each perspective: financial, ranked selection criteria, operations, sales, and marketing. This section will, again, be different for every data center and every data module design, but generally, the pros of a data module are a reduced OPEX, lower PUE, increased sellable units, more granular deployment of capital, freed up square footage, more efficient design, higher customer per rack density, incremental deployment, consumption scaled cooling, integrated DCIM, monitoring and control systems, faster provisioning, consistency with Tier certification and concurrent maintainability standards, sustainability messaging, and association with the latest data center technology. Data modules will also significantly reduce the amount of CAPEX vs. a traditional raised floor design because of the reduction of individual CRAC/CRAH units, and the minimized size and scale of the electrical, grounding, DCIM, fire detection, and suppression systems. The cons might be the need for staff training on the modules and the customers’ unfamiliarity with the technology. This section enables the reader to visually compare the designs from all relevant perspectives and selection criteria.
EIGHTH STEP: THE SUMMARY AND CONCLUSIONS
Display which option would be the most beneficial and explain why. Frequently, data modules provide many benefits alongside few detriments. Their high-density capabilities meet one of the most pressing demands on the data center industry. Greater densities also conserve physical space, which may be used for additional supportive equipment to increase power for sellable units. Regarding speed to market, modules allow for customer environments to be built rapidly, as needed, and in smaller increments than traditional raised floor environments. Incremental expansion shrinks capex investment and construction periods.
Data modules will likely reduce PUE, potentially allowing for the re-allocation of surplus power to the critical IT load; thereby increasing revenue or per rack kW. Data modules often come with built-in hot and cold aisle containment, which reduces power costs because it limits the amount of subsequent power needed to cool the heat produced by the critical IT equipment.
Data modules can reduce the number of cooling units needed, thus minimizing their resource provisioning, maintenance activities, and costs. Well-designed modules can also improve your monitoring and control systems, provide an additional layer of physical security, and, as a relatively new technology, be an attractive market differentiator. To end, summarize your conclusions concisely.
Multi-tenant data centers, like any business, always look for a market advantage. When the market demands certain requirements, such as greater densities and faster speed to market, it’s worth exploring how your business can meet those demands. Follow the above guide to craft a tailored TCO model and analysis of modular infrastructure vs. traditional raised floor and make the best choice for your data center’s future.