If you have followed my datacenter research and writing over the past decade, I have focused on X86, POWER, Arm, and accelerated infrastructure. Given the rise of cloud models and the reality that many mission-critical enterprise apps are staying on the mainframe, I felt it time I started to give the platform more attention. IBM’s Z mainframe infrastructure offerings, previously known as the “zSeries,” are an industry mainstay at this point, having been around since the 1960s, still powering environments that need the highest uptime, most security and largest transaction volumes.
IBM introduced the z900, its first 64-bit “Z” architecture offering, way back in 2000. The only way an offering can stay relevant that long is with constant improvement and evolution to keep up with the times. IBM’s older model, in many ways, disincented development of new apps and even growth of new apps on the platform. IBM has changed pricing models to meet many generations of client needs.:
- 1970 – 1999, IBM clients wanted simplicity and all you can eat plans at “Full Capacity”
- 1999 – 2018, clients asked for more variability to pay only for what was being used in what is called a “Sub-Capacity” (aka R4HA model)
With the rise of the cloud consumption and more non-mainframe options, some enterprise customers are moving some of their workloads off the mainframe and here we are today, with IBM introducing a new model it expects to last the next twenty years.
Last month, IBM announced its new Tailored Fit Pricing program for Z mainframes, a major pricing overhaul which I believe will make Z more accessible to enterprise customers and to better position it, in IBM’s words, as “the center of a secured hybrid cloud strategy.” We have written about many of the new consumption-based models and are excited about IBM’s as it removes a big objection to keeping and deploying new apps to Z. Let’s dive in and take a closer look.
Consumption-based model for software
The concept behind Tailored Fit Pricing is fairly simple: Z software pricing adjusts according to usage. Tailor Fit Pricing includes two different models—the Enterprise Consumption Solution and Enterprise Capacity Solution. The models look to deliver the flexibility of consumption-based pricing or “all-in” with the economies of scale necessary for IBM z/OS workloads. A huge benefit of these solutions is that they eliminate the need for capping, which is restrictive and can adversely affect responsiveness and service level availability. I believe that both models present good alternatives to the traditional sub-capacity “rolling four-hour average” (R4HA) pricing model, which tends to result in IT staff having to micromanage infrastructure to minimize costs. Additionally, these Tailored Fit Pricing models should make billing more predictable for businesses.
Enterprise Consumption Model
With the Enterprise Consumption Solution, the customer pays a monthly license charge and makes a baseline MSU (million service unit) commitment. Pricing is then based on usage, which is determined by measuring MSU consumption (which is aggregated hourly). To combat seasonal variability, this model includes built-in annual entitlements and reconciliation processes. IBM also touts this model’s aggressive growth pricing for MSU consumption beyond the committed baseline. This model should appeal to customers looking for a more flexible approach.
Enterprise Capacity Solution
The Enterprise Capacity Solution, on the other hand, is a full-capacity licensing model that determines pricing based on the estimated mix of workloads running while referencing the size of the physical environment. This should allow for simple, predictable budgeting, with a fixed monthly cost that IBM says covers the full capacity of the enterprise. On top of that, environments can be reconfigured within the full capacity, without changing the price or licensing requirements. Overall, this model should appeal most to enterprises who expect their workloads to expand and value cost predictability and operational simplicity.
What about the hardware?
While this pricing model applies only to software, there are a few caveats. With the Enterprise Capacity Solution, IBM prices the hardware, software, and maintenance all-in to the client, meaning hardware is included in the net price. To be clear, this is optional for IBM clients wanting to go further than the announced model. The announced model is software only, meaning clients already on z14 machines can choose this model without signing a new hardware contract. Beyond that, there’s an agreed-upon table that delineates the all-in price for additional capacity, although its not part of this announced model. On the other hand, software is totally independent with the Enterprise Consumption Model, with IBM operating under the assumption that the customer has its hardware capacity covered. In the event that they do not, they can turn to IBM’s “Capacity on Demand” offering, which they can toggle on and off for additional bursts of capacity on an as-needed basis.
Case studies and testimonials
IBM highlighted several testimonials from early customers taking advantage of Tailored Fit Pricing. Dillard’s, the national department store chain, says Tailored Fit Pricing is well-suited for its retail business where a disproportionate amount of the company’s sales volume transpires over the holidays (or in periodic clearance events). The client says Tailored Fit Pricing allows it to scale its IT resources to meet technology demands during these busy times, and in a cost-effective manner. You can read the Dillard’s case study here. One could see where this pricing model would be a boon to pretty much any business operating in the retail sphere.
IBM also highlighted Fiducia & GAD, who provides IT services to over 900 banks in Germany and has utilized Tailored Fit Pricing to better provision for unpredictable workload peaks. You can read the Fiducia & GAD case study here.
Overall, I think Tailored Fit Pricing is a smart move for IBM. I believe these flexible, cloud-like consumption models that could do much to decrease the slope of clients moving apps off the mainframe. Furthermore, I think it may motivate clients to move more apps onto Z where the highest levels of security, availability and transaction throughput are paramount. Tailored Fit Pricing is very interesting because most x86 consumption models are exactly the opposite—they meter the hardware and not the software. IBM is slowly but surely removing objections about the mainframe (pricing, chassis form factor, power, cooling, Linux) and raises the bar with security features and certain performance vectors. I am very interested to see where this goes on the application layer as Broadcom, Compuware and BMC have pledged support.
Note: Moor Insights & Strategy writers and editors may have contributed to this article.