Inc.
AI’s biggest obstacle may no longer be chips or computing power, but the money needed to keep everything running.
Intel and Broadcom CEOs Warn AI’s Next Big Bottleneck Won’t Be Chips
Author: Moses Jeanfrancois
The chief executives of Intel and Broadcom are warning the AI industry that the main constraint in its development will be access to capital.
According to CNBC, at the DBS Asia Future Forward Conference in Singapore on Monday, Broadcom CEO Hock Tan said cash flow could be the bigger obstacle to AI expansion rather than shortages of infrastructure. As the AI industry continues to boom, CEOs are looking to invest further, but are also increasingly becoming wary of its volatility and future.
Intel CEO Lip-Bu Tan echoed the same concerns about financing, according to CNBC. The tech leader pointed to the rising cost of capital for AI tokens and whether companies should continue investing.
“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,” Lip-Bu Tan said.
AI costs are driven by physical infrastructure, as well as data processing for model training, and ongoing usage fees. In recent years, costs have skyrocketed due to expanded use. According to Bloomberg, Meta Platforms, Amazon, Microsoft, and Alphabet are expected to spend roughly $750 billion in capital expenditures for AI this year alone.
″[In the] last two years we have seen a series of bottlenecks that keep coming out, but then we overcome it,” Hock Tan said at the conference. “Memory shortage today might not be that way a year from now.”
Tan expanded on the issue, stating that the industry was experiencing supply bottlenecks over the past two years, but expects those constraints to ease as production capacity increases, according to CNBC.
Capital crunch
Executives are leaning heavily on AI, primarily in the name of efficiency. But now companies must reckon with token usage, spend, and its high bill cost.
According to a report from The Wall Street Journal, only 11 percent of nearly 400 businesses were able to accurately forecast AI spending. 39 percent of respondents were 11 to 25 percent away from their target margin, while 35 percent of businesses were 26 to 50 percent off from their AI spend forecast.
Major companies are continuing to blow through their AI budgets, like Uber, which exhausted its entire 2026 AI coding budget within four months. But this hasn’t stopped companies from buying AI.
Intel’s CEO has emphasized that companies need to demonstrate tangible returns from AI investments in order for the business model to expand accordingly. This would mean AI would have to prove improved efficiency and even more rapid growth.
Both executives, though, remain hopeful for near-term demand for AI computing infrastructure.
“There is more demand for compute infrastructure than we can supply,” Hock Tan said. According to CNBC, the CEO added that the imbalance could continue through 2029 and possibly into 2030.
Credits: TCA, LLC.