Amazon (AMZN) has raised its 2026 capital expenditure forecast to approximately $220 billion, an increase from its earlier guidance of $200 billion. The company attributed this upward revision to escalating memory and component costs, alongside sustained investments in artificial intelligence and cloud infrastructure. The surge in AI-related demand has placed broad pressure on the semiconductor supply chain, tightening availability of high-bandwidth memory, advanced networking components, and specialized processors. Consequently, AMZN stock remains under close market scrutiny, as investors weigh the company's accelerating AWS revenue growth against its expanding capital requirements.

The revised spending projection was announced alongside Amazon's second-quarter financial results. The technology giant reported net sales increasing 20% year-over-year to reach $200.6 billion. Operating income saw a substantial 43% jump, climbing to $27.5 billion. For the third quarter, Amazon's management expects net sales between $197 billion and $202 billion, while operating income is projected to range from $22.5 billion to $26.5 billion.

A primary catalyst for Amazon's continued investment is the exceptional performance of Amazon Web Services (AWS). In the second quarter, AWS revenue surged 36.7% year-over-year to $42.2 billion. This represents the cloud division's fastest growth rate in 18 quarters. AWS achieved an annualized revenue run rate of $169 billion, fueled by demand for AI computing capabilities and cloud services. The unit's profitability also strengthened, with the AWS operating margin expanding to 39%—an increase of approximately 650 basis points compared to the same period last year. Furthermore, AWS reported a backlog of $496 billion, more than doubling its backlog from a year ago. Amazon confirmed that customers have already reserved much of its anticipated 2027 cloud capacity, with some demand extending into 2028.

To support these intensive AI workloads, Amazon is rapidly scaling its proprietary silicon development. The company's custom Trainium and Graviton chips have surpassed a $20 billion annual revenue run rate. Amazon is actively deploying its newer Trainium3 platform to serve large AWS customers. This push into in-house chip design positions Amazon to reduce its dependence on third-party GPU suppliers that have faced their own production constraints, while also offering customers a cost-competitive alternative for AI training workloads. Despite these expansions, the company acknowledged that customer demand continues to outpace available capacity. This persistent supply-demand imbalance has driven Amazon to accelerate the addition of data centers, servers, memory, networking equipment, and power resources.

Amazon's financial commitment reflects a broader trend of unprecedented capital expenditures among hyperscale cloud providers. Amazon's $220 billion plan exceeds Microsoft's anticipated 2026 capital spending of roughly $190 billion. Microsoft is working to scale its Azure infrastructure to capture growing demand, though power availability has limited some growth. Alphabet has also raised its 2026 capital expenditure forecast to a range of $195 billion to $205 billion, following an 82% year-over-year surge in Google Cloud revenue. Power availability and grid capacity have emerged as shared constraints across the industry, as new data center construction increasingly competes for limited electricity supplies. Collectively, the investments from these technology giants underscore that AI infrastructure expansion remains a central focus for the cloud computing industry.