IBM Mainframe Thrives in AI Era, Reports Record Revenue
IBM's mainframe systems continue to operate despite the prevalence of cloud computing, with the company's recent fourth-quarter earnings report indicating strong performance in its mainframe business.
Record Financial Performance
IBM CFO Jim Kavanaugh reported that the mainframe business achieved its highest fourth-quarter revenue in over two decades.
Revenue increased by 61% year-over-year, adjusted for currency, which contributed to a 17% rise in the infrastructure segment.
Enduring Market Dominance
Mainframe systems from IBM are utilized by industries requiring high reliability and security. Market Reports World indicates that 71% of Fortune 500 companies use mainframes. Approximately 92% of large financial institutions and 63% of government agencies employ mainframe systems.
IBM holds over 90% of the installed mainframe market.
More than 87% of global credit card transactions are processed on mainframe systems.
AI at the Core of Modern Mainframes
IBM has maintained mainframe relevance through continuous iteration, upgrades, and system improvements. The current z17 mainframe systems, which are contributing to recent revenue performance, are specifically designed for AI applications.
IBM's z17 mainframe system supports over 250 AI use cases, including:
- Loan risk mitigation
- Chatbot services
- Medical image analysis
- Retail crime prevention
The z17 can process up to 450 billion AI inferencing operations daily, a 50% increase from its predecessor, with average response times of one millisecond. IBM also offers the Spyre AI accelerator for installation in z17 mainframe systems, enabling the execution of more powerful AI models, such as IBM's Granite models.
Strategic Vision for AI Inferencing
IBM CEO Arvind Krishna stated during the fourth-quarter earnings call that he anticipates a shift in the AI inferencing market.
He projected that within three to five years, 50% of enterprise AI usage would occur in private clouds or on-premise data centers, with the remaining 50% utilizing public models.
This prediction suggests that AI inference in the cloud using frontier models is costly. Moving tasks with smaller, fine-tuned models and stable loads away from the cloud could provide cost savings for enterprises.
IBM typically releases a new mainframe system every 2.5 to 3 years. The z17, which shipped in mid-2025, is expected to be the primary mainframe offering until 2028. Given AI's strategic importance to IBM, future mainframe generations are likely to enhance AI inference performance.
Mainframes as Part of a Comprehensive AI Strategy
The mainframe is part of IBM's broader AI strategy, which also includes AI consulting services and the watsonx AI software platform. IBM's existing enterprise client base and mainframe install base provide an advantage for selling AI products and services.
IBM projects a total revenue growth of at least 5% in 2026, with software revenue anticipated to increase by 10%. Free cash flow is expected to rise by approximately $1 billion from the $14.7 billion reported in 2025. The mainframe business, alongside other AI ventures, is considered a significant factor in this positive outlook.
IBM states that the mainframe, while less central to the overall business than two decades prior, remains a "key contributor to IBM's profit and free cash flow."
The adaptation of the mainframe for the AI era has led to record sales for the segment. IBM is positioning itself as a significant player in the evolving AI industry.