Intel and Broadcom Say Money Could Become AI’s Biggest Bottleneck

Speaking at the DBS Asia Future Forward Conference in Singapore on October 12, Broadcom CEO Hock Tan argued that access to capital could become the most significant constraint on artificial intelligence infrastructure expansion.

The industry has spent years working through shortages of semiconductor wafers, high-bandwidth memory, and computing equipment. Tan suggested that financing may become an even more difficult problem as companies attempt to scale their operations.

Intel CEO Lip-Bu Tan expressed similar concerns at the conference, highlighting the cost of borrowing and the growing importance of demonstrating meaningful returns on AI investments.

Their remarks did not suggest that demand for AI computing is disappearing. Instead, they raised questions about how sustainably the industry can finance its next stage of growth.

From Chip Shortages to Capital Shortages

The early AI infrastructure boom was shaped by intense competition for advanced graphics processors. Companies raced to purchase Nvidia hardware, reserve manufacturing capacity, and expand data center operations.

Memory chips, networking equipment, electricity, and cooling systems also became major constraints.

Those physical limitations have not disappeared. But expanding AI capacity requires something beyond components: sustained financial investment.

Even a company that can purchase the necessary chips may struggle to finance an entire data center campus. That makes the availability and cost of capital central to future expansion decisions.

The AI Infrastructure Boom Needs Enormous Amounts of Financing

According to a Morgan Stanley estimate cited by Reuters, AI infrastructure could require approximately $1.5 trillion in external financing by 2028.

That estimate illustrates the extraordinary scale of investment required to build and operate the next generation of computing infrastructure.

At the same time, debt markets are showing signs of greater caution. Financial Times reporting indicates that AI-related debt issuance slowed substantially in September 2026.

The decline does not necessarily mean that AI demand is collapsing. Financing activity can fluctuate for several reasons, including interest rates, investor risk appetite, and the timing of large transactions.

Nevertheless, it demonstrates that the financial side of AI expansion deserves as much attention as the technology itself.

Why Higher Borrowing Costs Could Slow AI Growth

Building an AI data center often requires major upfront spending long before the facility generates revenue.

Companies must secure land, construction, processors, networking systems, electricity, cooling equipment, and ongoing technical support.

When borrowing becomes more expensive, the financial return required to justify a project increases.

A data center that appeared attractive under lower financing costs may become less compelling when interest expenses rise.

That could cause companies to delay, resize, or reconsider infrastructure projects even when demand for AI services remains strong.

The Billion-Dollar Question: Will AI Generate Enough Revenue?

Intel’s Lip-Bu Tan emphasized another important challenge: investors will eventually want evidence that massive AI expenditures are generating sustainable returns.

AI companies are already earning revenue through subscriptions, enterprise software, cloud computing, and developer APIs.

Broadcom’s Hock Tan pointed to companies such as OpenAI and Anthropic as examples of developers that have demonstrated meaningful revenue from advanced AI products.

But revenue growth and profitability are not identical. The cost of training models, operating inference systems, and expanding infrastructure can remain substantial.

The pressure to demonstrate results could encourage companies to prioritize projects with clearer customer demand and stronger economics.

Who Could Be Affected by an AI Capital Crunch?

The effects would depend on the severity and duration of any financing constraints. Companies with strong cash flow may have greater flexibility than firms relying heavily on external investment.

Could the Funding Squeeze Actually Improve AI Efficiency?

Financial pressure does not automatically mean technological progress will stop.

In fact, higher costs could increase demand for more efficient processors, smaller AI models, optimized software, and better utilization of existing hardware.

Cloud providers may focus more heavily on reducing the cost of each AI-generated response. Developers may prioritize systems that deliver practical value without requiring the largest available models for every task.

Such improvements could make AI services more economically sustainable. However, efficiency gains alone cannot eliminate every financing challenge, especially for extremely large infrastructure projects.

AI’s Future Depends on Economics as Much as Engineering

The artificial intelligence industry has demonstrated remarkable advances in model capabilities, coding assistants, scientific tools, and automated workflows.

But every major breakthrough requires infrastructure capable of delivering it reliably to users.

As the scale of investment grows, financial markets become increasingly important to the pace of development.

The warnings from Intel and Broadcom suggest that the industry’s next challenge may be proving that its technological ambitions can support a sustainable business model.