International Business Machines (IBM -1.05%) has shifted its focus to two areas during its decade-long transformation: hybrid cloud computing and artificial intelligence. The company has built a platform for each. Red Hat's ubiquitous software takes center stage in the hybrid cloud arena, while the recently launched watsonx platform now leads the charge in AI.
IBM has divested plenty of businesses over the past decade as part of its turnaround, but it's kept one of its oldest. The tech titan has been manufacturing mainframes since the 1950s. While the mainframe segment isn't as critical to IBM as it once was, the hulking systems are still widely used in certain industries, and each mainframe system sold generates additional revenue from software, services, and financing.
At the end of 2022, 45 of the world's top 50 banks used IBM's mainframe systems. Once every two and half years or so, IBM launches a new lineup of mainframes. The latest iteration, the z16, was announced in April 2022 and became available the following month.
The z16 came with a slew of innovations, including an integrated AI accelerator capable of running AI inferencing workloads in real time. This allowed financial institutions to use AI to analyze transactions on the fly, reducing the risk of fraud. A single z16 system could process 300 billion inference requests per day with one millisecond of latency.
A new mainframe is coming
Given the typical cadence of mainframe launches, IBM is likely to reveal its next mainframe system in the third or fourth quarter of 2024. A mainframe launch generally triggers an upgrade cycle, boosting revenue in the infrastructure segment substantially.
In Q3 of 2022, the first full quarter of z16 availability, sales of mainframe systems soared 98% year over year. This pushed total infrastructure revenue up 23% year over year, and it likely drove some incremental sales of software and services as well.
IBM's mainframe systems work well with the company's hybrid cloud platform, allowing customers to modernize their infrastructures while sticking with mainframes. One of the benefits of an IBM mainframe is longevity. An application written decades ago in COBOL, an ancient programming language that still powers many critical systems, will run on today's mainframes without issue.
The heavy usage of code written in COBOL is a problem for companies that are dependent on that code. COBOL isn't really taught anymore, so the pool of software engineers capable of maintaining those applications is shrinking. This situation is bad news for IBM because it could compel some customers to move away from mainframes.
IBM has recently come up with a solution that should help. The company launched watsonx Code Assistant for Z in October, a generative AI-powered assistant that is trained specifically to translate COBOL applications into modern Java code. This gives customers a way to modernize their software without giving up on the mainframe, and it gives IBM an additional way to generate revenue from its mainframe customer base.
Steady growth for IBM
After years of waffling between revenue growth and revenue declines, IBM is now in a good position to grow consistently, barring a severe economic downturn. The company is settling into its role of providing platforms, software, and consulting services to enterprises looking to modernize their IT infrastructures and adopt artificial intelligence technology. The mainframe itself isn't a growth business like hybrid cloud and AI, but it plays a role in both.
IBM expects to increase revenue by 3% to 5% this year excluding the impact of currency. A mainframe launch sometime next year will boost sales in the second half of 2024, but there are a lot of moving parts, and the company likely won't issue any guidance for 2024 until its next earnings report in January. The state of the economy is a wild card that could swing IBM's revenue by a few percentage points either way.
While a new mainframe will provide a temporary bump to IBM's top line, it's the company's efforts to reposition itself for the cloud and AI era that will drive growth in the long run.
