At the DBS Asia Future Forward Conference in Singapore on October 12, 2026, Intel CEO Lip-Bu Tan and Broadcom CEO Hock Tan warned that access to capital and rising capital costs could become the artificial intelligence industry's primary bottleneck. Although compute demand continues to outstrip chip supply—with Intel meeting only 50% of CPU orders—executives noted that high model training costs and uncertain long-term returns pose growing financial risks.
Capital constraints on AI growth
- ▪Intel CEO Lip-Bu Tan stated that alongside capital access and capital costs, a long-term challenge for the artificial intelligence industry is proving return on investment to investors
- ▪Broadcom CEO Hock Tan stated that access to capital could become a larger obstacle to artificial intelligence expansion than shortages of semiconductors, memory chips, or electricity
- ▪Intel CEO Lip-Bu Tan and Broadcom CEO Hock Tan warned at the DBS Asia Future Forward Conference in Singapore on October 12, 2026, that capital access and the rising cost of capital could become the biggest constraint on the artificial intelligence sector
Financial viability of AI labs
- ▪Broadcom CEO Hock Tan stated that OpenAI Group PBC is currently losing money because of high expenditures on training large language models
- ▪Anthropic PBC generated gross margins exceeding 80% excluding model training costs and revenue shared with partners, putting it on track for a second consecutive quarter of positive adjusted operating income in 2026, according to the Financial Times
- ▪Broadcom CEO Hock Tan noted that a sizable portion of enterprise revenue for artificial intelligence labs currently comes from pilot projects rather than sustained return on inference spending
Broadcom outlook on chip supply and demand
- ▪Broadcom CEO Hock Tan warned that computing infrastructure demand could slow if adoption of advanced artificial intelligence models disappoints or if model capability improvements begin to plateau
- ▪Broadcom CEO Hock Tan predicted that semiconductor wafer shortages would ease by 2028 or 2029, but the overall imbalance between computing demand and industry capacity could persist into 2030
Chinese open-source models
- ▪Intel CEO Lip-Bu Tan stated at the Singapore financial conference on October 12, 2026, that open-source language models from Chinese labs are proving good enough for budget-conscious users and driving efficiency in artificial intelligence
- ▪Open-source language models from Chinese laboratories gained a significant amount of traction over the past year
Intel quantum computing developments
- ▪Intel deployed a 12-qubit quantum computer developed using standard chipmaking equipment at the Argonne National Laboratory in January 2026
- ▪Intel CEO Lip-Bu Tan predicted at the October 12, 2026, Singapore event that quantum computing could become mainstream within three to five years
Debatable claims
- ▪Current enterprise AI revenue reflects temporary pilot spending rather than sustainable ROI
- ▪Capital access is a bigger constraint on AI growth than chip shortages
Story comments
Loading comments…