By 
SINGAPORE — Broadcom Chief Executive Hock Tan has warned that access to capital could become the biggest constraint on artificial intelligence infrastructure expansion, even as technology companies work through shortages of advanced chips and memory components.
Speaking at a panel in Singapore on October 12, 2026, Tan said the company was overcoming wafer and memory supply limitations but viewed financing as a more significant challenge ahead. His comments highlight a growing issue for the technology sector: building the infrastructure needed to support advanced AI systems requires substantial investment, and the cost of expansion continues to rise.
Broadcom supplies semiconductor and networking technologies used in data centres and AI computing infrastructure. Its position gives the company direct exposure to the growing demand for systems that can train and operate increasingly capable AI models.
AI Infrastructure Requires More Than Chips
The expansion of AI services depends on a broad network of infrastructure, including processors, high-speed networking equipment, memory, electricity, cooling systems and data centre capacity.
Technology companies are committing significant resources to these components as they compete to deliver faster and more capable AI services. While chip availability has attracted considerable attention, financing the construction and operation of large computing facilities presents a separate challenge.
Investments must often be made well before infrastructure generates enough revenue to recover its costs. Companies must also account for electricity expenses, equipment depreciation and the possibility that computing requirements or customer demand could change.
Revenue Potential Must Justify Investment
Tan pointed to the ability of leading AI developers, including OpenAI and Anthropic, to generate significant revenue from newer AI models and personal assistant products.
That commercial progress has strengthened expectations that AI services can support a large infrastructure market. However, revenue growth alone does not guarantee that every investment in computing capacity will produce an attractive return.
Infrastructure providers and their customers must consider how quickly demand will translate into recurring income, whether computing resources will remain fully utilised and how long current equipment will remain competitive.
For investors, the key issue is whether spending on AI infrastructure can be supported by sustainable business models rather than expectations of future growth alone.
Financing Costs Add Pressure
Large infrastructure projects are particularly sensitive to borrowing costs because they require substantial upfront spending and may take years to deliver their full economic benefits.
Higher interest rates can make debt financing more expensive and increase the returns investors expect from new projects. Companies that rely heavily on external funding may face additional pressure if revenue growth slows or capital markets become less receptive to large commitments.
The funding challenge also extends across the AI supply chain. Semiconductor designers, cloud providers and data centre operators all need to coordinate investments, making the sector increasingly dependent on the financial strength of its major participants.
What Comes Next for Technology Companies
Broadcom’s warning suggests that the next phase of AI expansion may be shaped as much by financial discipline as by engineering progress.
Technology companies will need to balance investment in new capacity against demand forecasts, operating expenses and the ability of customers to pay for increasingly sophisticated AI services.
If funding becomes more difficult to secure, some projects could be delayed or scaled back even if demand for computing remains strong. Conversely, companies with reliable cash flow and disciplined investment strategies may gain an advantage.
As the AI industry continues to expand, access to capital is likely to remain a key factor determining which infrastructure projects move forward and which businesses can sustain long-term growth.
Comments