If you are an IT professional who works on IBM System Z mainframes, I’m sure you’ve heard about zIIPs and other “specialty processors.” But unless you are mired in the day-to-day details of capacity planning, you may not really know what they are, what they do, and why they exist.
Beginning in the early 2000s, IBM started to introduce several different types of specialty processors. Basically, a specialty processor augments the main general-purpose CPUs by running specific types of workloads. So instead of running everything on the general-purpose CPUs, some workloads are sent to specialty processors to be run.
Why is this interesting? First of all, specialty processor workload is not subject to IBM or (for the most part) independent software vendor (ISV) licensing charges. So, running workload on a specialty processor can reduce cost.
Another benefit of specialty processors is that they are significantly cheaper to acquire than general purpose processors. A standard mainframe CP can cost more than half a million dollars, but the list price of a specialty processor can be significantly less than that. Furthermore, specialty processors can be purchased for a one-time charge per engine, including no-charge replacement by faster zIIP engines when upgrading to a new machine. Yet again, this means reducing costs.
But specialty processors can only run certain types of workloads.
What is a Specialty Processor?
There are three different types of mainframe specialty processors:
- ICF (Internal Coupling Facility),
- IFL (Integrated Facility for Linux),
- zIIP (Integrated Information Processor).
The first two are straightforward, the Internal Coupling Facility is used for processing coupling facility cycles in a data sharing environment; and the Integrated Facility for Linux is used for processing Linux on System Z workload on an IBM mainframe.
The zIIP, or Integrated Information Processor, is a bit more interesting. It is used for processing certain, specific types of distributed workloads. When you activate zIIP processors, some percentage of relevant workload can be redirected off of the general processors to run on the zIIP. Remember, the biggest benefit of redirecting work to the zIIP is that IBM does not impose software charges on workloads that run on the zIIP.
You probably noticed the phrase “relevant workload” in the previous paragraph. Not everything can run on the zIIP, only workloads that IBM deems as “new” are permitted. Originally, the zIIP was designed to support redirecting newer Db2 functionality, but over time the list zIIP-eligible workloads has grown. At a high level, the current zIIP-supported workloads include Java application programs, IBM z/OS Container Extensions (zCX), IBM Watson Machine Learning for z/OS, IBM Z System Recovery Boost, and for some types of Db2 for z/OS processing (e.g., XML, distributed queries, parallel queries, AI functions, and some aspects of IBM Db2 utilities). Other ISVs (e.g. BMC, Precisely, etc.) have zIIP-enabled their products so portions of those workloads can run on zIIP processors, too.
There are limits on the number of zIIPs that can be installed, up to two zIIPs per general purpose CPU. Second, IBM’s license agreement places restrictions on the kind of code that is eligible to run on a zIIP; the code must run in a z/OS enclave under the control of an SRB (service request block).
Additionally, not all zIIP-eligible workloads will actually run on the zIIP. It can be troublesome to understand exactly what is being redirected, exactly when, and exactly how much of the workload is being redirected. Nevertheless, the primary intent of the zIIP is to reduce your IBM software charge, and the more workload that can redirected to the zIIP, the more your monthly cost savings can be.
Are Specialty Processors Still Needed?
So, what is the future of the IBM mainframe specialty processor? The need for specialty processors is as strong or stronger than it has ever been. According to a recent BMC Mainframe Survey, 86% of the largest mainframe shops expect MIPS to grow in the coming year. So, with the potential of specialty processors to mitigate cost growth as capacity increases, it stands to reason that organizations should continue to utilize them.
Of course, you could make the case that reducing prices could be a more efficient way to mitigate the cost of mainframe software. When you think about it, the zIIP is really nothing more than a re-purposed general-purpose CPU, with controls that enable only certain types of workload to run on it. The only real difference is that a zIIP always runs at full capacity, even if the general purpose CPU is a sub-capacity model designed to run at a lower capacity.
But this is a simplistic way to look at the situation. Specialty processors are designed to encourage specific workloads – typically newer ones – to run cheaply on the mainframe. This enables IBM to reduce the cost of mainframe computing and encourages growing the footprint of what runs on the platform. Simply lowering prices would not likely server the same purpose.
By redirecting modern workloads (where there is much competition) to specialty processors, IBM can decrease the price of “modern” workloads while protecting the revenue it garners from “legacy” workloads, such as CICS and IMS transactions and batch programs (for which there is little to no competition).
An additional consideration is the cost of the specialty processors, which are much lower than the cost of a standard CP. And that means that specialty processors can reduce both software and hardware costs!
zIIPing Into the Future
So, what does the future hold for specialty processors? Although I don’t claim to be a soothsayer, the future looks bright. The first thing to acknowledge is that many organizations have implemented specialty processors and rely on them. If IBM were to consider eliminating specialty processors, it would negate the purpose for which they were developed. And it stands to reason that an alternate way of reducing the cost of newer workloads would need to be devised to help extend the viability of the IBM System Z mainframe.
In other words, eliminating specialty processors would be difficult unless IBM also created another way to achieve the same purpose. As long as specialty processors continue to work and are being used by customers, it is safe to assume that they will be viable for the long term.
On the other hand, introducing additional specialty processors might be more practical. For example, as use cases and computing techniques expand for Artificial Intelligence and Machine Learning, it stands to reason that IBM could introduce an AI specialty processor. Although it should be noted that many existing AI and ML capabilities are already zIIP-eligible and introducing a new specialty processor would be needless confusing. Security and cryptography are other areas that might benefit from specialty processors. But, of course, this is pure speculation on my part.
You should also keep your eyes on IBM’s pricing policies and announcements. As pricing metrics change, the cost patterns and expectations of users change. Therefore, a future pricing option might eliminate the need for specialty processors or, indeed, change the use cases for which they are designed.
For example, you might consider the impact if your organization changes from a sub-capacity pricing metric like AWLC to the newer tailored-fit pricing metric. Although tailored-fit pricing is not exactly a full capacity metric, it is not based on an R4HA peak. So, unlike sub-capacity pricing metrics, where only reduced workload during peak periods saves money, any zIIP redirection from general-purpose CPUs to zIIPs will likely save you money, eventually, when you use tailored-fit pricing.
The Bottom Line
It seems to me that the future is bright for specialty processors. There should be no worry about implementing them today to reduce costs. Of course, it makes sense to keep an eye on IBM’s announcements and pricing options to ensure that you are using specialty processors optimally at your site.