Post by Augustine Chiagozie (@pabloexchange)

๐Ÿ’ฐ Goldman Sachs maps where $7.6 trillion in AI spending is going

Goldman Sachs expects AI capital spending to reach $765 billion in 2026 and grow to $1.6 trillion annually by 2031, totaling $7.6 trillion over six years.

The breakdown is $5.1T for compute, $2.1T for data centers, and $358B for power.
โ €
The report assumes Nvidia captures 75% of compute spending, making it the biggest direct beneficiary of the AI buildout. At the same time, AI data centers are becoming far more demanding, with rack power rising from 5-15 kW in traditional facilities to 500+ kW in next-generation AI factories.
โ €
That shift also benefits infrastructure companies. Vertiv is positioned around liquid cooling and power systems, while Vistra stands to gain from rising demand for long-term nuclear power contracts needed to support AI data centers.
โ €
Goldman's conclusion is straightforward: chips, data centers, and power are the three layers every AI deployment depends on, and capital is flowing into all of them simultaneously.

๐Ÿ’ฐ Goldman Sachs maps where $7.6 trillion in AI spending is going Goldman Sachs expects AI capital...

0 likes ยท 0 comments ยท 0 shares