Investors Love AI, as Long as You’re a Cloud Host: Amazon’s $173 Billion Bet on the Future


Source: Russell Brandom / techcrunch.com

Amazon’s Impressive Q2 Earnings

Amazon reported better-than-expected second-quarter earnings on Thursday, sending the company’s stock up nearly 10% in after-hours trading. Net sales rose 20%, with cloud revenue standing out as a particular bright spot.

Cloud revenue, which includes sales from Amazon Web Services (AWS), rose 37% year over year to $42 billion for the quarter. This growth is a key driver of Amazon’s increasing spending on data centers and infrastructure.

Despite conventional wisdom that investors want companies to rein in data center spending, Amazon is continuing to invest heavily in its infrastructure. The company spent $173 billion on property and equipment for the fiscal year ended June 30, up from $107.65 billion in the previous year.

This spending includes investments in GPUs, natural gas turbines, and plots of land, as well as the development of its Trainium TPU and Arm-based Graviton processor chips. These chips are designed to improve the efficiency and performance of Amazon’s cloud services and will have a meaningful impact on the company’s margins.

Amazon’s decision to continue investing in its infrastructure, despite the potential risks, is a testament to the company’s confidence in the growing demand for cloud services. The company’s AWS and Amazon Bedrock businesses are expected to continue growing, with AWS revenue projected to reach $220 billion in 2026, up from the previous forecast of $200 billion.

However, not all companies are experiencing the same level of success. Companies like Meta, which have significant capex and no clear revenue source, are still experiencing intense skepticism from investors. Meta’s stock fell 8% after reporting quarterly earnings this week, as investors focused on its cash flow crunch and continued spending.

The contrast between Amazon’s success and Meta’s struggles highlights the complexities of the AI economy. While investors are increasingly focused on cloud hosts like Amazon, they remain skeptical about the underlying economics of AI labs and startups. This dynamic is a key driver of the growing competition and differentiation at every level of the stack, from cloud hosting to AI chip development.

In the end, the question of whether there is enough demand to justify the buildout of the AI stack remains unanswered. While Amazon’s hosting revenue may be someone else’s AI bill, the company’s success is ultimately tied to the demand for AI services. If demand doesn’t hold up, it will be a bad time for everyone involved.

As David Cahn’s $3 trillion question suggests, there are still many unknowns in the AI economy. However, one thing is clear: Amazon’s bet on the future of cloud services is a significant one, and the company’s investors are eager to see if it will pay off.

The company’s AWS business is expected to continue growing, with revenue projected to reach $220 billion in 2026. This growth is a key driver of Amazon’s increasing spending on data centers and infrastructure.

Amazon’s decision to continue investing in its infrastructure, despite the potential risks, is a testament to the company’s confidence in the growing demand for cloud services. The company’s AWS and Amazon Bedrock businesses are expected to continue growing, with AWS revenue projected to reach $220 billion in 2026, up from the previous forecast of $200 billion.