Premise

Hyperconverged public cloud IT services with a strong virtualized software stack have improved the value of applications to the lines of business, compared with traditional IT. Many vendors are bringing the benefits of public cloud services to on-premises, hyperconverged, private cloud solutions.  Wikibon has developed methodologies that focus on evaluating the business value of IT, and has applied them to 100s of enterprises over the last decade. This Wikibon research focuses on applying this methodology to private cloud deployments, and comparing it to traditional infrastructure. The premise is that this approach will deliver much greater “above-the-line” line of business value benefits compared with “below-the-line” IT operational cost savings.

Executive Summary

Digital businesses employ data as a strategic asset to better serve customers – and attack competitors. Consequently, cycle times for change are accelerating. Speed and agility are essential in all phases of business. This includes application management, which must accelerate delivery to sustain fitness between business activity and application capability. But it also includes infrastructure management, which must employ a lifecycle approach to ensure that infrastructure resource pools can be quickly, easily, reliably, and stably reconfigured and re-targeted to better serve application and business needs.

The absolute requirement to sustain application fitness in a digitally transforming world ties infrastructure to business value in unambiguous – and unprecedented – ways. The right infrastructure, coupled with the right infrastructure lifecycle management approach, can facilitate Agile development, shave weeks off application time-to-value, and dramatically reduce the risks and failures of change.

To model the nature and scale of this business value, Wikibon compared a hyperconverged Dell EMC VxRack SDDC against a “do-it-yourself” (DIY) white box approach over three years. Our model presumed a large enterprise: 10,000 employees and revenues of about $3.4 billion. The average number of active users online at any one time is assumed to be 1,000, as the end users are doing many other tasks in addition to using the applications. The maximum number of active users is assumed to be 2,000, and the equipment sized to meet that demand. To fully test business returns, we modeled both “below the line” (IT cost savings) and “above the line” (business value generated) benefits in the comparison.

Figure 1. Highlights of Hyperconverged Technology Business Value research project

The below-the-line IT cost savings are $1.7M, but the above-the-line business value from earlier implementation, faster updates, better availability and recovery, better security, and improved compliance is $11.6 million (see Figure 1). The total value of this approach is $13.3 million.

Wikibon research thus finds the above-the-line value components (the value to the business) are dominant, and over seven (7) times higher than the below-the-line benefits. It is important that above-the-line benefits are evaluated carefully when deciding on long-term adoption of hybrid cloud infrastructure.

Hyperconverged Reference Architecture

Prior Wikibon research concluded that Dell EMC was the leading vendor for True Private Cloud, which Wikibon defines as an edge or on-premises platform that truly operates like a public cloud service. Wikibon research also determined VMware and Nutanix to be the leading virtualized hyperconverged platforms. For this research Wikibon chose a reference architecture based on the Dell EMC VxRack SDDC hyperconverged infrastructure which utilizes the hyperconverged VMware private cloud deploying vSAN and NSX. It is compared to a traditional “best-of-breed” infrastructure approach that deploys a “White Box” solution with traditional separate SAN and networking components.

Measuring Hyperconverged Private Cloud Value

While faster time-to-value is a key business benefit of public cloud, the lower cost of public cloud infrastructure may be the chief business benefit for highly seasonal or intermittently used applications with high compute usage (e.g., some machine learning applications).

There are many good business reasons for running applications on-premises, either centrally or at the Edge, including avoiding moving data, which is very slow and expensive, and instead moving cloud services and application code to the data. In addition, there are often compelling legal and business reasons for keeping data on-premises. Large enterprise data centers have many applications which are dependent on data from other applications, and where the data originates from a local source. As a result, almost all large enterprises are following a hybrid cloud strategy, where data movement and application execution is optimized across multiple public and private clouds.

The key values of private cloud on-premises for production application systems compared with a traditional approach include:

  • Below-the-Line Cloud Requirements (benefits accrue to the IT Budget)
    • The cost of IT equipment and software to implement and run the applications.
    • The cost of staff to install, operate, and provide recovery and compliance services for IT applications.
  • Above-the-Line Cloud Requirements (benefits accrue to the Line-of-Business Budget)
    • Improved time-to-value for new IT application projects.
    • Improved time-to-value for application maintenance; Improved application and data security.
    • Improved application availability and recoverability.
    • Improved capability and flexibility to meet internal, industry and state compliance requirements for applications and data.
    • Improved flexibility to enable moving applications to the data or data to the applications across a common hybrid cloud infrastructure (this benefit was not measured in this exercise).

Hyperconverged VMware

VMware is the de-facto standard for virtualizing infrastructure in many enterprises. VMware has expanded to provide the compute, storage, and networking virtualization, as well as support for containers and Serverless computing. They have embraced a multi-cloud and hybrid cloud enterprise strategy by providing a common technology and management environment spanning multiple private and public clouds.

VMware Cloud Foundation is the core software suite of the Dell EMC VxRack SDDC. The VMware Cloud Foundation SDDC Manager is the management and orchestration software. VMware and Dell EMC have combined the VMware SDDC software into a hyperconverged integrated solution with VMware Cloud Foundation being fully integrated into the Dell EMC VxRack SDDC which includes:

  • VMware vSphere
  • VMware NSX
  • VMware vSAN
  • VMware SDDC Manager

The Dell EMC VxRack SDDC VMware vRealize Suite option is included for monitoring of cloud and IaaS operations.

The VMware software is assumed to be already installed at the customer site, and is not included in the cost evaluation.

This report builds on previous Business Value of Technology (BVT) research by Wikibon and extends it to look at the above-the-line business benefits that are available to the enterprise. Table 1 below shows the detailed line items, assumptions, and the calculations used to derive the benefits shown in Figure 1 above.

Hyperconverged Dell EMC VxRack SDDC

Dell EMC VxRack is a tested and validated rack-scale hyperconverged infrastructure built on Dell PowerEdge servers and Dell EMC Vision Intelligence Operations software with integrated networking and automated software lifecycle management. Direct-attached storage is virtualized by a combination of VMware vSAN and Dell software to provide a shared pool of block storage, similar to SAN storage. The network is virtualized by a combination of physical Cisco Nexus switches and VMware NSX. Today, the VxRack SDDC is the most advanced platform for a VMware multi-cloud environment.

The white box reference system for this research is based on a standardized architecture that includes x86 nodes, top of rack, and rack interconnect switching, as well as pre-loaded VMware software components. The Dell EMC VxRack SDDC addresses the compute and storage needs, as well as requirements for networking (physical and virtual), cloud operations, and system management automation.

The configuration used for this research includes 24 VxRail nodes in a single rack. Each physical rack is pre-integrated with physical network switches. These control network traffic, redundancy, and management for out-of-band connectivity. This configuration can be expanded to multiple racks, with additional network equipment for east-west traffic.

Hyperconverged Business Case Analysis

Hyperconverged Business Case

The reference model is a large enterprise with 10,000 employees supporting a revenue of $2.4 billion. Our model pegs an average salary of $52,000 for those 10,000 employees, with the fully burdened salary being 65% higher or $85,800. Wikibon research finds that on average, about 10% of an employee’s time is spent on tasks where they are actively using IT. Lastly, we assume that application value is a function of the total average contribution of an employee to the revenue generated by the organization, i.e. 4 times fully burdened salary, and that it is generally going to be the same whether they are using the application or doing other business tasks. Therefore, the total application value is calculated as $85,800 x 1,000 x 4 = $343 million/year. This is the top line of Table 1 below.

We presume our modeled business faces the same digital business transformation pressures as any real business. The basis of any digital business is data: digital businesses employ data as a strategic asset, traditional businesses do not. Using this simple-to-say, hard-to-apply concept gives real meaning to the notion of digital business transformation. A business is further along in a digital business transformation if it is further along in applying data to restructure its engagement model, re-institutionalize its value streams, and reorganize its people and processes.

The drive to digital business transformation is accelerating application changes. Some of this acceleration is endogenous: using data to empirically identify and prioritize opportunity and iteratively pursue them shortens decision and application cycles. Some of the acceleration is exogenous: competitors are doing the same, which constantly alters the portfolio of options available to a business. A digital business transformation, then, causes profound drift in application fitness. Simply put, in today’s digital business environment, application value exhibits an annual decline because playing fields are constantly reshaping in response to competitive, regulatory, and customer experimentation.

For this model, Wikibon presumed that applications decline in value by a conservative 10%/year. Work is required to maintain the application and infrastructure environment to ensure the business value of the application and the productivity of the users. In a traditional white box environment it is assumed that these updates and upgrades are added once a year.

Wikibon’s research also shows that one of the greatest sources of friction constraining efforts to sustain application fitness is infrastructure lifecycle management. More regular changes to application capabilities catalyzes more frequent, and frequently more profound, changes to infrastructure supporting these applications. Superior infrastructure lifecycle management enables an enterprise to more easily pool crucial infrastructure resources, thereby streamlining and simplifying the complex tasks of allocating critical infrastructure resources to match application priorities. It’s this streamlining of infrastructure lifecycle tasks that can accelerate time-to-value of changes that improve application fitness. It’s this simplifying that diminishes failure rates and risks associated with these changes.

To sustain application fitness (and accelerate adoption of infrastructure advances), the hyperconverged Dell EMC VxRack SDDC approach allows a fully integrated upgrade to take place every quarter. Why? Multiple reasons, but at the top of the list are three:

  • SDDC integration with VxRack reduces the administrative costs of adding scale-out capacity to nearly zero.
  • Dell EMC’s life cycle management (LCM) approach to pre-tested engineering and automated distribution of updates reduces administrator tasks and minimizes failure rates; and
  • Integrated networking ensures that system balance is easily sustained.

In comparison, the Traditional White Box solution requires enterprise IT to pull together all the updates required, and to test it for themselves to ensure everything is working correctly. The cost and elapsed time to complete these tasks results in the average VMware infrastructure and system environment for the Traditional White Box solution being upgraded about once per year.

Table 1 calculates this loss of application value in both the traditional white box environment and the hyperconverged environment. The value lost to application decline in the Dell EMC VxRack SDDC case is $1.3 million better than the white box case in the first year (Line 2 of Table 1 and greater detail in Table 4).

The difference in time to implement the infrastructure can also be important. In this analysis Wikibon assumes that the infrastructure is being implemented to adopt new application software. Our research shows that the time-to-value advantage of an integrated Dell EMC Hyperconverged VxRack approach, relative to typical “do it yourself” options, is 14 days. This additional value is $2.1 million in the first year vs. the white box approach (see Table 1, line 3 and details in Table 4).

Table 1 (Line 4-6) also shows the reductions in the number of expected days lost due to unavailability to end-users and recovery time, due to security breaches, and due to compliance. In any one year, there may be no losses, but a larger loss in a subsequent year will result in a significant impact over a 3- year period. The detailed assumptions and calculations are shown in Figure 4 below.

The total benefit in above-the-line benefits to the lines of business in the first year is $5.2 million (Line 7), and $11.6 million over 3 years (Line 8).  The differences in IT costs are derived from Table 2 below. It shows the $1.65 million net savings to the IT budget, a below-the-line saving.

Table 1 shows the value of using a hyperconverged system in terms of both increased application value (above-the-line) and decreased cost (below-the-line) to be a total of $13.3 million over 3 years. The net present value is $12.1 million (discounted at 5% over 3 years), as shown in Table 3.

Hyperconverged Below-the-Line 3-year Business Case

Table 2 show the differences in IT costs between the hyperconverged Dell EMC VxRack SDDC vs. a traditional White Box DIY approach. The main benefits of the Dell EMC VxRack SDDC are lower IT personnel costs for implementation, operational support and system administration. The Dell EMC VxRack SDDC equipment costs are estimated to be about 10% higher than a traditional white box systems.

These figures are used in Table 4 in the Footnotes section to calculate the below-the-line benefits to IT.

Hyperconverged Above and Below-the-Line 3-year Business Case

Table 3 below shows the business case for both above-the-line and below-the-line items, and shows the total business case for implementing the infrastructure. It takes the bottom line figures from Table 4 in the footnotes, and spreads the costs and benefits over three years.

It shows that the total 3-year business benefits of implementing the application are $761 million using a Dell EMC VxRack SDDC approach, and $748 million deploying a traditional white box DIY approach.

The analysis shows an overall benefit of $13.2 million to deploy a hyperconverged Dell EMC VxRack SDDC. The net present value is $12.1 million (discounted at 5% over 3 years), as shown in Table 3 below.

Conclusions

The key to realizing the IT and business value that “True Private Cloud” installations can provide is to ensure they are simple to set-up and run. All aspects of support need to be automated, and remote operation is a prerequisite. Fully integrated hyperconverged stack systems such as the Dell EMC VxRack SDDC have changed the game by lowering operational costs relative to bespoke systems based on white boxes. This is the “below the line” benefit, the direct savings in IT cost.

However, when dealing with strategic decisions on hybrid cloud, the business value of a VxRack hyperconverged solution shows up clearly when the above-the-line business benefits are considered. After all, the reason for implementing IT projects is to improve the productivity, improve the revenue, and reduce the risks to the business that IT supports.

VMware and Dell EMC have helped IT run more efficiently and effectively by integrating the hardware and software, and by enabling servers to run compute, networking, storage and management in a tightly integrated system. Wikibon believes this is an essential step to allow flexibility of application deployment, and to provide a common infrastructure and management software and procedures across both private and public cloud deployments. In addition, the migration costs for establishing this environment are minimal for established VMware users.

Wikibon believes on-premises hyperconverged infrastructure together with an integrated virtualization software stack is a business imperative for creating True Private Cloud environments competitive with the public cloud.

Action Item. Wikibon believes methodologies that estimate both the above-the-line business value and below-the-line IT operations value to the business should be used to evaluate hyperconverged private cloud solutions.  Wikibon strongly recommends that the hyperconverged solutions such as the Dell EMC VxRack SDDC be included in every RFP for large-scale VMware infrastructure deployments.

Appendix: Hyperconverged Business Value Detailed Methodology

Table 4 below shows all the detailed assumptions and calculations in the Wikibon Business Value Methodology. Two environments are compared. The first is a mixed workload running a traditional VMware “White Box” deployment with a traditional server, SAN storage and networking. The second is the same VMware mixed workload running on a hyperconverged stack solution. The reference model is the Dell EMC VxRack SDDC solution. The source for our business assumptions is a body of Wikibon Business Value Research we have compiled and updated over the past decade as well as more recent Business Value Research on Hyperconverged Systems.

Figure 1 provides a detailed graphical summary of the data in Table 4.

The assumptions and detailed calculation of each line of Table 4 are as follows:

  • Line aa gives the assumption of the average number of employees who are actively engaged in using the IT system. This number (10%) is based on detailed Wikibon CIO research on calculating application value.
  • Line a shows an average of 10% (1,000) of all the employees using mixed workload applications running on a VMware system at any one time. The expected average to peak ratio is 1:2, and the peak number of users running applications is approximately 2,000.
  • Line b shows an average salary of $52,000 for the 1,000 full time equivalent (FTE) staff.
  • Line c shows a typical overhead for benefits, space, and overall management of 65%.
  • Lined calculates the average cost to the business for an active application user. The calculation is d = b × (1 + c).
  • Line e calculates the cost of active application users in a year. The calculation is e = d x a. The fully burdened cost of the 1,000 FTEs of $8.6 million/year.
  • Line f shows the revenue contribution multiplier assumption for active users while they are using the applications. The underlying assumption is that users are as productive while using the system as the average of the other tasks they do as part of their job (e.g., meetings, customer discussions, etc.). Of course, this figure can vary considerably according to the industry with a range of 2 to higher than 20. For this analysis, we use an average of 4.
  • Line g shows the average revenue contribution of an application user (FTE) to the business of $343,000 per year. The calculation is g = f × d.
  • Line h shows the total revenue contribution from applications being run by users for a year. The formula is h = g × a.
  • Line i shows the depreciation in application business value that occurs every year, if no maintenance is applied. The assumption is 10% as an average over many VMware applications. As a comparison point, enterprise application providers (ISVs) usually charge about 20% for software maintenance and support
  • Line j shows the number of infrastructure system updates that can achieved per year. The hyperconverged Dell EMC VxRack SDDC solution can achieve 4 updates per year. These updates are pre- tested by Dell EMC, and include the latest version of the VMware software together with all the hardware updates (e.g., microcode changes) required. They are delivered once every quarter. In comparison, the Traditional White Box solution requires enterprise IT to pull together all the updates required, and to test it for themselves to ensure everything is working correctly. The cost and elapsed time to complete these tasks results in the average VMware infrastructure and system environment being upgraded about once per year.
  • Line k shows the percentage of application updates that require systems update/year.
  • Line l calculates the lost application value/year from application depreciation. The calculation formula is l = h × i × 50% × k ÷ j. If the application value loss over a year from depreciation is 10%, the average application value loss over the whole year will be 5%, or 50% of 10%. The 50% in the formula translates loss at the end of the year to average loss. This figure varies significantly by vertical industry.
  • Line m is the number of years in the business case (3).
  • Line n is the percentage of new applications introduced every year. The assumption is 8%.
  • Line o is the time to achieve initial value from a new application. The time to install and be operational for end-users.
  • Line p is the days lost/year because of unavailability and time to recover. This represents the business value of productivity loss from application unavailability. This does not include the impact to the business brand or consequential loss, as the application workload is mixed running on VMware. Other methods would be used for high-value systems of record.
  • Line q is similar to line p for security breaches.
  • Line r is similar to line p for compliance failures.
  • Line s calculates the loss of business value from delayed time-to-deploy in the first year. By being able to bring up new application systems and updates faster, the business benefits are accrued faster. The formula is s = h × n × o ÷ 365. This benefit is realized in just year 1. Line t calculates the loss of business value from unavailability and recovery time over three years. Although all vendors claim very high levels of availability for the hardware, the largest amount of time is spent in the recovery of the application and any lost data. The number of incidents of failure is reduced by more frequent updates to the system. The calculation reflects the loss of time for end-users. The potentially much more significant losses of value (e.g., revenue loss, brand image loss, etc.) are not included in our calculation. The formula is t = h × m × p ÷ 365
  • Line u calculates the loss of business value from security breaches over three years. Although all vendors claim high security, every day there are many new security threats. Again, the largest amount of time is spent in the recovery from the breach and any compromised data. The number of incidents of security breaches is reduced by keeping the system more up-to-date. The calculation reflects the loss of time for end-users. The potentially much more significant losses of value (e.g., revenue loss, brand image loss, etc.) are not included. The formula is u = h × m × q ÷ 365.
  • Line v calculates the loss of business value from compliance failures over three years. The number of incidents of compliance failure is reduced by keeping the system more up-to-date. The calculation reflects the loss of time for end-users. The broader losses of value (e.g., revenue loss, brand image loss, etc.) are not included. The formula is v = h × m × r ÷ 365.
  • Line w calculates the total loss of business value in the first year. The formula is w = l + s + t + u + v
  • Line x calculates the sum of the application value losses over 3 years. The formula is x = ((s + t + u + v) × m) + l. The last column calculates the difference between the hyperconverged approach and the traditional white box approach, which is $11.6 million.
  • Line y calculates the total application value generated by the VMware stack. The formula is y = h × m – x.
  • Line z shows the Below-the-Line IT costs for equipment and support for both environments. The details of the IT equipment costs, maintenance and support are shown in Table 3 in the footnotes below. Savings of
  • $1.7 million in the IT budget is shown in the final column.
  • Line ab calculates the cost of the users running the application. The formula is ab = f × m.
  • Line ac calculates the total below-the-line costs of running the VMware systems, including IT costs and end-user costs. The formula is ac = z + ab.
  • Line ad calculates the net application value created by the VMware stack. The formula is ad = y – ac.