Broadcom Inc. is in talks to raise as much as $80 billion in new debt to fund a chip financing arrangement, one of the largest corporate borrowing efforts tied to the artificial intelligence buildout, according to reports on Friday. The figure would rank among the biggest debt packages ever assembled by a technology company.
Early accounts placed the target above $60 billion, but later reporting from CNBC’s David Faber put the range at $70 billion to $80 billion, citing people familiar with the discussions. The wide gap reflects how quickly the terms are moving as banks and investors weigh the scale of demand for AI infrastructure spending.
The financing would support Broadcom’s expanding work designing custom chips for the largest operators of data centers, a business that has become central to its growth story. Demand for specialized processors has climbed sharply as companies race to build the computing capacity needed to train and run large AI models.
Behind the numbers, the deal fits a wider pattern of capital flooding into the sector. It follows moves such as Nvidia’s roughly $500 billion financing arrangement with Wall Street partners and Alphabet’s multibillion-dollar bond offering, both aimed at funding the infrastructure underpinning AI services.
For investors, the borrowing carries clear stakes. A debt load of this size raises Broadcom’s interest costs and ties its balance sheet to continued growth in AI orders. Should demand cool, the company would still owe repayment on financing raised during the current boom.
Traders have positioned around the stock accordingly. Jeff Kilburg, chief executive of KKM Financial, described the company as “primed for a big gain” while advising strategies that protect against downside risk, reflecting the mix of optimism and caution surrounding AI-linked names.
Broadcom has ridden the AI wave through much of the past two years, reporting strong bookings from a small group of large customers building out data centers. The company designs application-specific chips alongside its networking and infrastructure software divisions.
For consumers and businesses relying on cloud services, deals of this scale help determine how fast new AI capacity comes online and, eventually, what it costs to access. The financing being arranged now shapes the computing power available over the next several years.