12 October 2026
Two chip bosses say money, not chips, may be what slows the AI build-out
Broadcom's Hock Tan and Intel's Lip-Bu Tan both said in Singapore that the cost of financing could become AI's tightest constraint, even as demand still beats supply.
For two years the story of the AI boom has been about shortages: not enough advanced chips, not enough memory, not enough electricity. On Monday the heads of two of America's biggest chipmakers pointed at a different scarce ingredient: money.
What happened
Speaking at a DBS conference in Singapore, Broadcom chief executive Hock Tan argued that raising money could prove a bigger obstacle to AI growth than chips, memory or power. "I think might be the biggest constraint is capital," he said.
Intel CEO Lip-Bu Tan (no relation) made a similar point. He flagged the rising cost of capital, meaning the interest rate or return investors demand before they hand over cash, and added that over the longer run investors will want proof that AI spending pays off.
Neither executive sounded gloomy about the next few years. Hock Tan said demand for computing infrastructure is still running ahead of what the industry can supply, and that the gap could last through 2029 and perhaps into 2030. Lip-Bu Tan said he did not expect a major slowdown in the next two or three years.
Hock Tan also argued that the physical bottlenecks tend to get solved. He suggested today's memory crunch may look different a year from now, that wafer shortages (wafers are the silicon discs chips are cut from) could ease by 2028 or 2029, and that power supply should improve as new plants come online.
Why it matters
Data centers are expensive and slow to pay back. When bond yields rise, the borrowing that funds new buildings and chip orders gets pricier, and a project that looked profitable at one interest rate can look shaky at a higher one. If the people selling the chips are raising that concern, it is a sign the financing side of the boom is getting more attention.
Both CEOs tied the long-run outlook to whether businesses actually adopt AI and see real gains from it. Hock Tan said demand could cool if advanced models disappoint or their improvements level off; Lip-Bu Tan said companies will increasingly have to show concrete returns, through efficiency or new lines of business.
For scale, Broadcom was worth roughly $1.8 trillion and Intel about $567 billion at the October 7 close, according to Noorult data. See the Broadcom and Intel company pages for their full financials.
What to watch
- Bond yields, which feed directly into what it costs to fund new data centers.
- Whether big AI spenders start pointing to measurable returns in their earnings reports.
- Signs that memory and wafer shortages are easing on the timeline Hock Tan described.
The companies in this story
Based on reporting by CNBC: Intel, Broadcom CEOs warn capital crunch could become AI's biggest bottleneck. Facts, not advice.
More news
- Anthropic's IPO, in numbers: 825 million shares at $70 to $78, and a $1.4 trillion price tag
- Trump says the Nvidia scandal didn't endanger US security, and stands by Lutnick
- Nvidia's first earnings without Huang: what Kress will be asked on Tuesday
- Buffett gives away $5.5 billion of Berkshire stock in his Thanksgiving letter
- Banks are using AI to trim back offices. Proof that it adds revenue is harder to find
- IMF chief says the world economy leans too hard on AI as circular deals top $1 trillion