SKN | Goldman Sachs Maps a $7.6 Trillion AI Infrastructure Buildout as Supply Constraints Shift

Key Takeaways:

• Goldman Sachs projects approximately $7.6 trillion in AI infrastructure investment through 2031, highlighting the scale of the capital cycle now developing around artificial intelligence.
• The five largest hyperscalers are expected to spend approximately $800 billion in capital expenditures in 2026, with annual spending potentially reaching $1.4 trillion by 2028.
• Goldman identifies memory, networking and power infrastructure as increasingly important constraints as AI clusters scale, rather than GPUs alone.
• Memory producers are already reporting gross margins of roughly 80%, indicating significant pricing power created by constrained supply.

Goldman Sachs is framing artificial intelligence as a multi-year infrastructure investment cycle rather than a conventional technology spending boom. Its latest projections point to approximately $7.6 trillion of AI infrastructure investment through 2031, placing banks such as Goldman at the center of a capital-intensive transformation spanning computing, memory, networking and energy infrastructure.

Goldman Sachs Sees AI Capital Spending Entering a New Scale

The five largest hyperscalers are expected to spend approximately $800 billion on capital expenditures in 2026, representing a 94% increase from 2025. Goldman’s analysis suggests the absolute level of spending remains the more important consideration as the growth rate begins to moderate from exceptionally high levels.

For Goldman Sachs, this represents a significant expansion in the financing and advisory environment surrounding AI infrastructure. Capital requirements at this scale can create demand across investment banking, financing markets and institutional capital as companies continue building the physical infrastructure required for increasingly intensive AI workloads.

The Infrastructure Bottleneck Is Moving Beyond GPUs

Goldman’s analysis points to a changing constraint within the AI buildout. As clusters grow into projects requiring hundreds of billions of dollars in annual spending, the limiting factors increasingly include memory, networking capacity and power infrastructure.

This matters for the financial architecture of the AI cycle. The infrastructure supporting AI systems is becoming more interconnected, meaning capital expenditure is no longer concentrated solely around processors. The broader buildout requires increasingly large investments across the physical systems that allow those processors to operate at scale.

Supply Constraints Are Creating Pricing Power

One of Goldman’s clearest signals comes from the memory market. Supply constraints have pushed memory producers’ gross margins to approximately 80%, more than twice their historical average.

For Goldman Sachs, this demonstrates how infrastructure bottlenecks can redistribute economics within the AI investment chain. When supply cannot expand as quickly as demand, producers with constrained capacity can gain substantial pricing power. That dynamic can influence capital allocation throughout the broader infrastructure ecosystem.

What Goldman’s Forecast Means for the Bank

The $7.6 trillion projection places Goldman Sachs in an environment where capital formation becomes as important as technological innovation. Hyperscalers require enormous funding capacity, infrastructure companies require capital to expand supply, and investors need financing and advisory services to participate in the resulting investment cycle.

For Goldman, the strategic opportunity therefore extends beyond following individual AI companies. The bank can participate across the financing architecture supporting data centers, power systems, networking, semiconductors and other infrastructure required by the AI economy.

The key issue for sophisticated investors is ultimately the durability of this capital cycle. Goldman’s figures indicate that spending is expected to remain enormous even as growth rates normalize. That makes the quality, financing and productivity of AI infrastructure investment increasingly important variables for the bank and its institutional clients.

For a confidential discussion regarding your cross-border banking structure, technology-sector exposure or international wealth strategy, contact our senior advisory team.