IBM

The company's history is a series of reinventions, and understanding that pattern is the key to understanding what IBM is. It was formed in 1911 as the Computing-Tabulating-Recording Company, a merger of firms that made scales, time clocks, and punch-card tabulating machines. Under Thomas J. Watson Sr., who took over in 1914 and renamed the firm International Business Machines in 1924, it became the dominant supplier of punch-card data processing equipment, the closest thing the pre-electronic world had to computing. In the 1950s and 1960s, under Thomas Watson Jr., IBM bet the company on the System/360 mainframe, a single family of compatible computers that could scale from small to large. The bet succeeded and defined commercial computing for a generation. For decades IBM was the most valuable and most feared technology company on earth, so dominant that it spent years fighting a federal antitrust case.

The story since then has been one of repeated near-death and renewal. IBM nearly collapsed in the early 1990s when the personal computer, a product it had helped popularize, undermined the economics of its mainframe franchise. Lou Gerstner arrived as chief executive in 1993 and chose not to break the company apart, instead reorienting it toward services and integration, selling not just machines but the work of making technology function for a business. That pivot toward services and software, and away from commodity hardware, has guided IBM ever since. The company sold its personal computer division to Lenovo in 2005 and later shed chip manufacturing, printers, and other lower-margin businesses, concentrating on software, consulting, and the mission-critical systems that large institutions cannot easily replace.

Today IBM organizes itself into four reporting segments. Software is the largest and most profitable, and it includes Red Hat, the open-source company IBM acquired in 2019 for about 34 billion dollars, along with automation tools, data and AI platforms, and transaction processing software that runs on mainframes. Consulting is the services arm, a large professional organization that helps clients design, build, and run technology projects, increasingly centered on cloud migration and AI adoption. Infrastructure is the hardware business, dominated by the IBM Z mainframe line and the Power family of servers, along with storage. Financing is a smaller segment that provides credit to support purchases of IBM technology. The economic shape of the company is that Software and Consulting generate the bulk of revenue and most of the recurring, higher-margin income, while Infrastructure delivers large but cyclical revenue tied to the timing of new mainframe releases.

The durable core of the business is easy to overlook because it is unglamorous. The IBM mainframe, now in the z17 generation built on the Telum II processor, still processes an enormous share of the world's commercial transactions, including a large portion of credit card payments, bank transactions, and airline reservations. Mainframes are expensive and decades old as a concept, yet they persist because the software written for them is deeply embedded in the operations of the institutions that depend on it, and rewriting that software to run elsewhere is risky, slow, and costly. This creates one of the strongest competitive moats in technology, a switching cost so high that customers keep buying new generations of hardware rather than face the danger of migrating away. Mainframe revenue is lumpy, rising sharply when a new model ships and tapering between cycles, but the installed base is remarkably loyal, and each hardware refresh pulls along years of associated software and support revenue.

The Red Hat acquisition is the second pillar of the modern strategy and the foundation of IBM's hybrid cloud positioning. Red Hat sells commercially supported open-source software, most importantly the Linux operating system and OpenShift, a platform for running applications in software containers across many environments. IBM's argument is that most large enterprises will not move everything to a single public cloud. Instead they will keep some systems in their own data centers, spread other workloads across Amazon Web Services, Microsoft Azure, and Google Cloud, and need a common layer that lets software run consistently everywhere. Red Hat provides that layer, and it does so without requiring customers to commit to IBM's own cloud. This is a deliberate position. Rather than trying to beat the dominant cloud providers at their own game, where IBM trails badly, the company sells the tools and services that help enterprises manage a mixed environment, taking a neutral role that competitors tied to a single cloud cannot easily occupy.

The third pillar is artificial intelligence, delivered through watsonx, the platform IBM launched in 2023. The watsonx family includes tools for building and tuning AI models, a data store for the information that feeds them, and governance software for monitoring and controlling how models behave. IBM has aimed this squarely at corporate buyers rather than consumers, emphasizing data privacy, the ability to run models inside a company's own environment, and integration with the enterprise software those companies already use. More recently the company has pushed into agentic AI, software agents that can carry out multistep tasks, with watsonx Orchestrate connecting to enterprise applications from vendors such as Salesforce, Microsoft, ServiceNow, Oracle, Adobe, and SAP. The strategy ties AI tightly to consulting and to the existing software estate, so that IBM sells not just models but the work of putting them into production inside complex organizations.

A defining event of the recent era was the separation of Kyndryl. In November 2021 IBM spun off its managed infrastructure services unit, the part of the business that ran clients' data centers and IT operations on a day-to-day basis, into an independent public company called Kyndryl, which trades under the ticker KD. The unit was large but low-margin and slow-growing, and shedding it allowed IBM to focus on higher-value software and consulting and to present a cleaner growth story. The move reshaped the company's reported revenue and remains central to how IBM frames its current trajectory. Around the same time the company has continued to acquire software and automation businesses, including HashiCorp, which it bought in early 2025 to strengthen its infrastructure automation portfolio.

Competition is intense and varies by segment. In consulting IBM competes with Accenture, the large Indian outsourcing firms such as Tata Consultancy Services, Infosys, and Wipro, and the consulting arms of the major accounting firms. In cloud and AI infrastructure it sits well behind Amazon Web Services, Microsoft, and Google in raw scale, which is why the hybrid and neutral positioning matters so much. In enterprise software it overlaps with Oracle, SAP, Microsoft, and a wide field of specialized vendors. In AI it faces both the cloud giants and a rising set of model developers. IBM's defense in most of these fights is the same: deep relationships with the largest and most conservative institutions, a reputation for security and reliability, and the gravitational pull of the mainframe and the Red Hat platform that keep those customers within reach.

Leadership sits with Arvind Krishna, who became chief executive in April 2020 and chairman in early 2021, and who also holds the title of president. Krishna is an engineer by training who spent his career inside IBM Research and the software business, and he was a principal architect of the Red Hat acquisition before taking the top job. His tenure has been defined by simplifying the company, completing the Kyndryl separation, and concentrating capital and attention on hybrid cloud and AI. The management style under Krishna has emphasized disciplined acquisitions of software companies, partnerships rather than head-on competition with the cloud leaders, and a steady return of cash to shareholders through a long-standing dividend that the company has prioritized maintaining.

Beyond the near-term strategy, IBM continues to invest in quantum computing, an area where it is among the recognized leaders. The company operates a fleet of quantum processors accessible over the cloud and has published a roadmap aiming for a large-scale, fault-tolerant quantum system, named Starling, around the end of the decade. Quantum is a long-horizon bet with little current revenue, but it reinforces IBM's identity as a research-driven company and could matter enormously if the technology matures. IBM Research, one of the largest industrial research organizations in the world, remains a distinctive asset, a source of both patents and credibility with the scientific and enterprise communities.

The risks are specific and worth naming. The mainframe moat is durable but not growing, and the entire franchise depends on customers continuing to choose not to migrate away, a choice that becomes incrementally easier as cloud tools mature. The consulting business is cyclical and exposed to corporate technology budgets, which contract in downturns. In cloud and AI, IBM is a follower competing against companies with far greater scale, faster growth, and deeper pockets, and its differentiated hybrid position depends on enterprises continuing to value neutrality over the convenience of a single vendor. The company carries meaningful debt from its acquisitions, and its long record of declining or flat revenue before the recent stabilization has made some observers cautious about how much of the AI opportunity it can actually capture. Execution risk runs through the entire strategy, because IBM must convert its installed base and its consulting relationships into durable software and AI revenue faster than its advantages erode.

The forward question for an observer is whether IBM has finally found a durable engine of growth or has simply assembled a stable, cash-generative portfolio around a slowly shrinking core. The bull case is that the combination of an irreplaceable mainframe franchise, a genuinely neutral hybrid cloud platform in Red Hat, and an enterprise-focused AI strategy gives the company a defensible and growing position selling to the world's most demanding customers. The bear case is that IBM remains a follower in the most important technology shift of the era, with a moat that protects but does not expand, and that its size and history make rapid reinvention difficult. What is not in dispute is that International Business Machines Corporation has survived more than a century of technological upheaval by repeatedly remaking itself, and that the current bet on hybrid cloud and AI is the latest chapter in that long pattern of reinvention. Whether this reinvention proves as successful as the System/360 or as halting as some of the company's other turns is the central thing to watch.