The chief executives of Intel and Broadcom have cautioned that access to capital may become a significant limitation on the artificial intelligence surge, even as demand for computing infrastructure keeps surpassing supply.

Broadcom CEO Hock Tan stated that financing may turn into a larger hurdle for AI growth than shortages of semiconductors, memory chips, or electricity, as the sector hurries to construct the infrastructure required to support ever more advanced models.

"[What] I think might be the biggest constraint is capital," Hock Tan remarked on Monday at the DBS Asia Future Forward Conference in Singapore.

Intel CEO Lip-Bu Tan shared similar worries about financing, highlighting the increasing expense of capital and doubts about whether firms making large AI investments will ultimately achieve adequate returns.

"If there's any challenge in the front, it's capital and the cost of capital, and then the other part is in the longer run, people want to see the ROI," he said.

Their comments point to a rising issue for the AI sector, which has invested massive amounts in data centers, cutting-edge chips, and power infrastructure in expectation of soaring demand for computing capacity.

These alerts arrive as climbing borrowing costs threaten to make financing more difficult for tech companies pouring money into AI infrastructure. Elevated bond yields could drive up the expense of funding data center construction and chip acquisitions, possibly straining the finances of projects that might take years to yield returns.

Hock Tan noted that the industry has faced repeated supply bottlenecks over the past two years, but he expects some of these limitations to diminish as production capacity grows.

"[In the] last two years we have seen a series of bottlenecks that keep coming out, but then we overcome it," he said. "Memory shortage today might not be that way a year from now."

He indicated that wafer shortages could lessen by 2028 or 2029, while power availability might also get better as new infrastructure is launched.

In spite of the financing worries, both executives stayed optimistic about short-term demand for AI computing infrastructure.

"There is more demand for compute infrastructure than we can supply," Hock Tan said, adding that the mismatch could last through 2029 and possibly into 2030.

Intel's Tan likewise said the industry was unlikely to see a major downturn in the next two to three years, pointing to expanding use of AI applications and demand for computing power.

But maintaining that growth beyond the next few years will hinge on how swiftly businesses embrace AI and whether the technology provides substantial economic advantages, the executives said.

Broadcom's Tan warned that demand for computing infrastructure might decelerate if the uptake of advanced AI models falls short or if enhancements in their capabilities start to level off.

Lip-Bu Tan, in the meantime, stressed that companies would increasingly need to show concrete returns from AI investments via better efficiency and novel business models.