Broadcom CEOs warn that capital could become a bigger obstacle to artificial intelligence expansion than shortages of chips, memory or electricity. Broadcom CEO Hock Tan and Intel CEO Lip-Bu Tan said financing and the cost of investment may limit how quickly the industry builds computing infrastructure.
Speaking Monday at the DBS Asia Future Forward Conference in Singapore, Hock Tan called capital a possible major constraint. Lip-Bu Tan pointed to both the rising cost of capital and the need for companies to show returns on their artificial intelligence spending.
Broadcom CEOs Warn About Financing
“[What] I think might be the biggest constraint is capital,” Hock Tan said. Lip-Bu Tan said the challenge includes “capital and the cost of capital,” followed by whether businesses can demonstrate a return on investment.
The source report describes rising borrowing costs as another pressure on companies financing AI infrastructure. Higher bond yields could make data center construction and chip purchases more expensive, putting pressure on projects that may take years to generate returns. The comments place financing alongside the supply limits that have shaped the infrastructure buildout. Companies have been investing in data centers, advanced chips and power infrastructure as demand for computing capacity grows. The executives’ warning comes as businesses race to build facilities and equipment for increasingly sophisticated models. Hock Tan said demand for compute infrastructure exceeds what the industry can supply. He said the sector has repeatedly faced bottlenecks, then overcome them, as capacity changed.
Some Supply Bottlenecks May Ease
Hock Tan said the industry has encountered successive supply bottlenecks over the past two years, but expects some constraints to ease as production capacity expands. He said memory shortages today might not persist a year from now.
He suggested wafer shortages could ease by 2028 or 2029. Power availability could also improve as new infrastructure comes online. Those possibilities address specific supply pressures, while the executives’ financing concern centers on whether companies can fund projects and eventually justify the expense.
Hock Tan said the imbalance between demand and available computing infrastructure could continue through 2029 and potentially into 2030. Lip-Bu Tan said the industry was unlikely to see a significant slowdown over the next two to three years, citing adoption of AI applications and demand for computing power.
Returns Will Shape Longer-Term Investment
The executives tied longer-term growth to business adoption and evidence that artificial intelligence produces meaningful economic benefits. CNBC reported that the executives made the remarks as firms continue to spend on facilities, chips and power in anticipation of more computing demand. The CNBC report on their warning details the financing concern and the supply outlook. Lip-Bu Tan said companies would increasingly need to show tangible returns through improved efficiency and new business models.
Hock Tan also cautioned that infrastructure demand could slow if adoption of advanced AI models disappoints or improvements in their capabilities plateau. That leaves the investment outlook dependent on both continued demand and the returns companies can demonstrate, even as near-term demand remains strong.
The financing question also connects to broader leadership coverage, including how CEOs frame organizational priorities. The discussion over infrastructure investment sits alongside reporting on how technology leaders describe AI. Readers can also see how entrepreneurs approach long-term growth in a profile of Malaysian business leaders.
The next test is whether companies adopting AI can translate that use into returns sufficient to sustain infrastructure spending. The executives’ remarks identify financing and demonstrable business value as factors that will shape the industry’s longer-term expansion.