Right-Sizing Solver for Enterprise IT

Introduction

Zypr is the FinOps tool for right-sizing data center infrastructure services hosted on-premise or off-premise using bare-metal hardware.

Using discrete-event simulation and dynamical system techniques, Zypr gives IT infrastructure service planners, architects and engineers the power to harness the predictive capabilities of simulation and gain the forward-looking visibility necessary to deliver services at market competitive rates — a foundational enabler to higher rates of digital innovation and business growth (see Jevons paradox).

Right-Sizing Infrastructure Services

Imagine using your valuable time to simply describe the characteristics of an infrastructure service and a few minutes later having a wholistic and synchronized view of hardware, software, power and space resources that will be required to satisfy projected service demand, while attaining your resource utilization goals.

Imagine not having to reconcile disjoined supply, demand and utilization forecasts for individual resources and eliminating the unwitting over-provisioning it produces. Zypr uniquely solves all of that for you — for whatever time horizon you prefer — in minutes.

Zypr empowers infrastructure service teams to explore, predict and shape the least-cost way to satisfy a service’s projected demand, sourcing planners to more precisely synchronize contractual commitments with wide-ranging lead-times, and financial planners to enable leaner, more agile inventory policies.

At the heart of Zypr is its simulation engine that models a service’s resource composition as an ongoing stateful process, not a single event. Every simulation automatically balances resource supply to service demand by generating future capacity add events, consistent with the utilization boundaries you specify, and simultaneously determines the required resources for each event.

With a wide assortment of objects to fully describe future events, evolution rules, costs, and constraints, Zypr provides infrastructure service teams with the tools necessary to test and validate the optimal pathway to cost-effectively right-size their services — and easily transform event-based, operational strategies into into calendar-based resource forecasts.

Imagine how much time and resources your team can save with a more complete and precise picture of your future resource needs.

Higher Resource Utilization

Imagine investing less money in hardware, software, power and space resources to satisfy the same level of service demand.

Maybe you’ve already invested in real-time workload placement tooling with similar expectations. If you have, Zypr will help you extract more value from your investment. If you haven't, that's okay. Either way, Zypr enables enterprises to do more with less because sustaining a lower spend rate from higher hardware utilization is fundamentally a supply-chain synchronization problem. And Zypr provides the information necessary to solve it.

Zypr uses a hardware utilization boundary constraint rule, which you set, to help govern the future evolution of a service's resource composition. The upper utilization boundary signals when a service requires additional capacity (i.e., an add capacity event), whereas the lower boundary limits how much capacity to add. Increasing the upper utilization rate delays a capacity add event. It doesn't change required resources (i.e., reduce spend) unless the lower boundary is also increased.

If higher average hardware utilization is accomplished solely by raising the lower utilization boundary, periodic capacity add events will occur at a faster cadence. Either strategy only produces cost savings if sourcing commitments are reduced, particularly software. Using Zypr’s predictive power provides service teams the forward-looking visibility necessary to make good short-term decisions consistent with the future outcomes they seek.

Zypr makes it easy to identify unnecessary capacity, or evaluate a hybrid strategy that steers costly peak demand to a public cloud. Whatever your strategy, Zypr is the tool for determining how to safely execute a leaner resource profile that delivers more for less.

Hardware Arbitrage

Imagine having more control over your software spend.

Infrastructure services are typically composed of different software stacks, whose costs are usually correlated to hardware performance and run 2-10 times more than hardware. This creates counter-intuitive economics, where the value of hardware exponentially increases as its share of total cost decreases.

This paradox makes optimally right-sizing hardware or software inseparable; one cannot be solved without solving the other. It also means there are countless ways to acquire and deploy hardware — as an alternative currency — to increase computing rates per software license, watt, and space at a lower total cost (i.e., arbitrage), rather than spending a lot more to add additional software licenses and consume more power and space resources.

Zypr gives enterprises the predictive power of simulation to simultaneously optimize a service's resource composition for whatever time horizon is required — and minimize financial commitments for software. Because these commitments precede those for hardware, right-sizing hardware is the complementary transaction that completes arbitrage monetization — which you purposefully created by right-sizing software.

The harsh reality of over-provisioning software is it diminishes the value of future hardware refresh; arbitrage opportunities vanish. Worse, it promotes longer refresh cycles and the misdiagnosis that little can be done to control software spend.

Zypr is the only FinOps tool that gives enterprises the power to predict upcoming arbitrage opportunities and deploy hardware to its highest economic value during its entire service-life — at the lowest total cost.

Zypr vs Traditional TCO

Commonly used TCO models were never designed to optimally solve the complex choices modern architectures afford, where workloads are portable, hardware is fungible, and resources are composable.

Zypr is a new type of FinOps tool to assist infrastructure planners, architects, and engineers to proactively identify, plan and deliver infrastructure services where hardware inventory is not treated as a fixed or static asset. Instead, hardware inventory is managed as a resource that can be purposefully organized and reorganized to its highest economic value, over its entire service-life, to deliver services at highly competitive rates.

Feature	Zypr	TCO
Identify server service-life	✓	✓
Identify server economic-life by resource pool	✓	×
Simulate the evolution of a resource pool's composition	✓	×
Find optimal right-sizing solutions and enable hardware arbitrage	✓	×
Synchronize supply-chain to sustain higher resource utilization	✓	×

When you give Zypr a try, you don't have to imagine savings. Zypr also lets you simulate right-sizing using a fixed-time refresh policy. Simply toggle between "Fixed" and "Optimal" simulation types, execute a few simulations to test your assumptions, and compare the results.

Scalable Solution

Zypr can be implemented and scaled in a variety of ways that makes getting started easy and future growth seamless.

• Select the type of simulation that best fits the characteristics of the pool being modeled. No need to use the same type of simulation for every pool.
• Get started simulating one or two resource pools that are particular pain-points or offer the biggest or fastest ROI.
• Use Zypr's API GUI to run simulations or write scripts for more complex simulations or simply to automate into current workflows.
• Select a shared simulation runtime environment or a dedicated environment for more control over simulation queue and run times.

How to Get Started

Zypr provides a rich set of objects to describe many types of resource pools and shape how future resources evolve over your preferred time horizon.

Once your registration is confirmed, you may access a repository that contains predefined "starter" models, and "how-to" examples, that introduce the various ways a pool may be described. You can read about those models here. You will also find the repository lookup codes to fetch complete model definitions by use-case using this endpoint.

Follow these steps to register and run your first simulation. You may use Zypr free for 30-days to evaluate how it can best deliver value to your organization. If you need additional time to evaluate Zypr for your organization, please reach out to us.