IBM A Dark Horse To Acquire Intel Foundry Services?

For decades, IBM was a vertically integrated device manufacturer, designing and fabricating its own chips for its pioneering mainframe and server businesses. The IBM Microelectronics Division, founded in 1966, was a hub of domestic semiconductor operations in Burlington, Vermont, and East Fishkill, New York. It was responsible for groundbreaking innovations, including the first mass-manufactured DRAM, which was used in its early computers. By the early 2000s, IBM began to transition its business model, attempting to become a contract foundry and sell its fabrication services to other companies. This strategic shift was in response to the escalating capital requirements of the semiconductor industry, which was becoming a volume business. In 2015, IBM sold its entire microelectronics division to GlobalFoundries for a negative price of $1.5 billion, offloading an unprofitable business but attempting to secure a long-term supply of advanced chips.

In my mind at least, the idea of IBM forging a deep strategic partnership with Intel’s nascent foundry business is not a reversal of its 2015 decision to exit semiconductor manufacturing but a necessary and forward-thinking evolution of its business model. My analysis of the situation suggests that this move is a critical maneuver designed to address a profound supply chain vulnerability, capitalize on an existing technical synergy, and strategically align with pressing U.S. national security and economic priorities. This is a story about two companies finding common ground in a very competitive and capital-intensive industry.

Going back to the 2015 divestiture to GlobalFoundries (I was at IBM at the time) confirms that it was a transaction of last resort. The original decision to sell its fab business for a cash payment of $1.5 billion was intended to offload an unprofitable, CapEx-intensive business while securing a long-term supply of advanced chips. That strategy failed spectacularly, culminating in a series of legal disputes that highlighted the risks of relying on an arms-length foundry relationship. The trust was eroded, and the core purpose of the deal—a guaranteed, long-term supply of chips for IBM's mission-critical products—was effectively nullified. The experience from that failed partnership provides the most compelling rationale for why a different approach is needed today. A simple contractual agreement is not enough in this market. IBM now works with Samsung for its Telum II and Power 11 chips.

My analysis of IBM's current state reveals a paradox. While the company has shed its general-purpose manufacturing assets, it remains a highly CapEx-intensive enterprise. The company is not CapEx-averse; it is simply more selective. IBM has redirected its investments toward high-margin, future-focused technologies such as quantum computing and the design of advanced processors for its mainframe and server businesses. For example, IBM announced a massive $20 billion investment over the next decade in the Hudson Valley region for research and manufacturing, and a pledge in 2025 to invest $150 billion in the U.S. over five years. These figures demonstrate that IBM is willing to make monumental capital investments, but only in areas it deems strategically vital and where it can maintain a leadership position. The company's R&D budget has been steadily increasing, a clear indication that focused R&D and strategic CapEx remain central to its long-term vision.

This brings me to Intel's IDM 2.0 strategy. It offers a compelling solution. With its ambitious process roadmap—including the breakthrough Intel 18A node featuring RibbonFET and PowerVia technologies. Intel is positioned as a powerful, domestic manufacturing partner that is hit hard times. The timing is ideal. The synergy between the two companies is both technical and geopolitical. IBM's in-house chip designers, who pioneered a 2-nanometer nanosheet technology, are working on the same foundational Gate-All-Around (GAA) architecture that underpins Intel's next-generation nodes. A partnership would accelerate the transition of IBM's research into volume production.

Furthermore, both companies are deeply tied to U.S. government and defense contracts and are major stakeholders in the CHIPS and Science Act. A formal alliance would create a unified, domestically aligned technology powerhouse, uniquely positioned to secure future government funding and contracts for mission-critical microelectronics. This is not just a business decision; it is a national security imperative. The creation of a fully domestic, end-to-end supply chain, from IBM's Albany research lab to Intel's U.S. fabs and finally to IBM's Poughkeepsie assembly facility, would be a formidable competitive advantage.

A strategic partnership would not only resolve IBM’s supply chain vulnerabilities but also create a modern hybrid business model where IBM's R&D prowess is fused with Intel's scalable, domestic manufacturing expertise. This would be the ultimate fulfillment of the original strategic intent behind the GlobalFoundries divestiture, correcting the mistakes of the past by forging a deeper, more integrated relationship. The cost of a minority equity investment and a long-term capacity reservation agreement would be a fraction of the cost of re-entering the foundry business, while providing a far more reliable and strategically aligned solution. This is a very compelling case for both companies to come together.