Skip to content

Paradoxical Economics of Enterprises

Use of Expensive Software Stacks

Mid-to-large scale enterprises typically operate many resource pools, with a range of software stacks and costs, that provide differentiated features and service levels to hosted applications. This results in pools with very different cost structures.

alt

The Paradox

When hardware cost is a fraction of a pool's total cost, right-sizing economics change and become paradoxical, such that:

As hardware's percentage of a pool's total cost decreases, its economic value increases with respect to software, power and space consumption when those costs are inversely correlated to hardware performance or efficiency.

The Counter-Intuitive Effect

The higher a pool's cost structure ratio (i.e., Kratio) the more important it is to shape hardware assignment; simply, there are more costs that can be displaced. The Kratio therefore acts as a coefficient to the relative gain in hardware performance when determining its economic value.

For example, a modest 20% gain in hardware performance for a pool with a Kratio of 10.0 is economically equivalent to a 50% gain for a pool with a Kratio of 2.0. The ideal uptake rate of better hardware, as well as its duration of use (i.e., hardware economic-life), will therefore vary over time and from one pool to another.

alt

Being able to prospectively differentiate economic value among different resource pools and hardware models — from newly available hardware versus existing inventory — is essential information in order to deploy hardware to its highest economic utility over its service life and thereby minimize total cost. Zypr enables this by:

  • Identifying and evaluating the many combinatorial ways by which a pool's resource compositional mix may evolve over time.
  • Distinguishing hardware retirements from inter-pool reassignments and pool recomposition (i.e., distinguishing hardware economic-life versus its service-life).
  • Accomplishing both at the lowest total cost with respect to the unique infrastructure economics of enterprise IT.

Ravello Analytics, LLC