CFOs: The mainframe isn’t a money pit. It’s time to see the value
May 17, 2022 | Best practices
Simon Ratcliffe Principal Consultant, Advisory
Establishing a modern IT strategy can be a hazardous journey for many organisations as there is the need to balance investment priorities, a fundamental aspect for all CFOs. So what’s the mainframe got to do with it? There is significant value that can be gained if CFOs take a closer look at the impact of mainframe and their organisations’ modernisation strategy. Nowhere is this balance more apparent than when looking at the mainframe estate within your organisation. Many IT strategies these days are predicated on out with the old and in with the new – ‘cloud first’ is the headline we most often read on investment plans.
We see this more often than ‘realising value from the mainframe’, but we must not overlook the intrinsic value of the mainframe. Legacy has become a pejorative phrase in technology terms, but the years of data stored in mainframes, their inherent ability to simply process efficiently and the integral nature of their technology means they are often more than old iron that needs to be replaced.
The value of the data alone is significant and with the current emphasis on the success of data driven organisations, it should not be overlooked. The finance function was an early adopter of data driven analysis and the success of this approach across other functions will be heavily influenced by the quality of the data that is already owned.
Mainframes are being constantly developed to make them more modern and relevant to today’s business requirements. Ally this knowledge with the treasure trove of data contained on them and businesses should think long and hard before simply ditching mainframe in favour of cloud. However, one of the issues with the mainframe is that it is difficult to compare it with the cloud.
For many years the cost of the mainframe has been calculated in terms of storage and MIPs. Storage makes sense and is one of the more sensitive costs in cloud environments, but MIPs is less straightforward. Calculating the Millions of Instructions per Second (MIPS) is, arguably, similar to calculating the vCPU and RAM requirements in cloud but there are significant differences that have to be taken into consideration.
The underlying mainframe hardware has changed beyond all recognition in the past 20 years. Not only has technology become faster, it has become smarter as chip technology has developed. Some of the next-generation IBM chips have 256 cores (compare this to an ultra-high-performance Xeon chip with 56 cores) which deliver incredible processing performance. The issue with this underlying performance is that it is only of value if the applications are aware of and can benefit from the hardware. Thus, the debate on selecting the right platform relies on placing the right software on the right hardware.
Where simple raw speed and reliability is required, mainframes are still in a league of their own with the capacity and speed that is built into modern mainframes. However, whilst they are ideal systems of record, mainframe architecture doesn’t readily lend itself to systems of engagement, such as mobile applications. The art of understanding and exploiting the value of the mainframe from the CFO perspective is understanding which platform yields the best performance for the investment made. In other words, some systems will undeniably perform better on a mainframe and some will be better suited to cloud deployment and the financial case will still make sense.
The problem in many organisations is that, when business cases are created, they are built to ‘prove’ that cloud is the solution and that mainframes are both holding the business back and adding to the ever-increasing technical debt. As with many reviews, when the desired outcome becomes one of singular intent such as ‘cloud first’, many of the benefits of retaining the mainframe are airbrushed out. We need to be very wary of taking this singular view as we will potentially miss the right answer in our haste to re-locate to the cloud.
Truly benefiting from mainframe
However, we also need to realise that to truly benefit from the mainframe, we will need to invest and modernise. Like any technology, simply ignoring it will result in progressively less value being realised and the mainframe will then become the anchor holding back the business. Modernising the software to ensure that it can maximise the benefit of the underlying hardware is crucial. Just as we would with any platform, we need to balance hardware and software performance but, when we achieve this, the financial results are impressive. For example a financial services organisation saved $10 million and gained 50% increased processing capacity from modernising their mainframe.
CFOs will see a positive return from a modernisation investment
As CFOs, our approach to investment is predicated on measuring returns. Many CFOs will see a positive return from a modernisation investment in their mainframe for specific services that they will never realise by simply shipping the mainframe estate to the cloud. However, if the mainframe remains unchanged and unloved then the financial projections are likely to be bleak as costs will spiral as performance degrade is relative to the platforms around the mainframe.
One of the arguments to retire the mainframe is technical debt, where the assertion is that the mainframe is a financial legacy that is simply draining current funds that could be better spent elsewhere. The issue with this approach is that it misses the point of technical debt. A system that has served well for many years does not represent technical debt, rather it is a sunk cost. Investing in that system is not increasing technical debt but rather making an investment with a defined return, exactly as organisations will in the cloud. The primary difference between cloud and mainframe by default is that the cloud is an operating expense and the mainframe is traditionally a capital expense, but that is no longer a distinction with Mainframe as a Service so that investment can now become an operating expense.
Mainframes can deliver great value and significant returns
The language surrounding the mainframe has created an image problem that has distorted the financial value of mainframes. With modernisation and investment, mainframes can deliver great value and significant returns on investment to organisations for many years to come.
One of the problems that often arises from the decision to migrate off the mainframe simply because it is legacy is that organisations actually increase their technical debt. As CFOs we need to take a fresh view of our mainframes and rather than simply asking ‘how do we migrate away’ we should ask ‘how can we derive more value’ from our mainframe. For too long we have ignored the mainframe or hoped it would fade away as we believed that it was a drain on our budgets. The reality is that mainframes are so integral to those organisations that use them that a modernisation approach is much more likely to yield financial return than attempting to migrate away.
If we examine the criticality of the systems we run on the mainframe and the value of the data created as we run these processes we will be able to understand the value of this environment. Modernising what already exists to allow modern, agile access to data and more modern operating systems to provide more options for development will often yield greater financial benefit than establishing a long project to migrate away to an alternate option. If the lack of appetite for capital expenditure is driving this behaviour, then Mainframe as a Service provides an Opex based alternative.