Recently, while Microsoft has reduced some of its leasing obligations in its data center expansion, both Amazon and Alphabet appear determined to advance with their plans. Despite Microsoft planning to invest approximately $80 billion in infrastructure capital expenditures for artificial intelligence (AI) this fiscal year, its fiscal year concludes in June, with some early-stage projects paused due to divergent needs between itself and AI partner OpenAI. OpenAI is independently expanding its capabilities and is part of Project Stargate, intending to invest $500 billion in AI data centers over the coming years.
In contrast, Amazon and Alphabet are preparing to make significant investments in 2025. Alphabet recently reaffirmed its intention to spend $75 billion on data center capital expenditures this year, while Amazon is set to spend around $100 billion, undeterred by potential tariff impacts.
Amazon CEO Andy Jassy, in a letter to shareholders, highlighted AI as a transformative force, advancing rapidly beyond previous technological innovations. Similarly, Alphabet CEO Sundar Pichai, at the Google Cloud Next ’25 conference in Las Vegas, described the potential of AI as incredibly significant.
Historical trends suggest that Amazon and Alphabet’s substantial investments could yield positive results. Amazon has a precedent of substantial capital expenditures, having built a comprehensive warehousing and logistics network to enhance the delivery speed of its products, which significantly contributed to its status as a leading e-commerce entity. Furthermore, Amazon played a pioneering role in the infrastructure-as-a-service industry through Amazon Web Services (AWS), now its most profitable segment.
Alphabet, too, endured initial losses with its Google Cloud business but recently observed a notable turning point in profitability within its Google Cloud segment, with a 142% increase in operating income to $2.1 billion in the last quarter.
In 2017, Goldman Sachs analysts identified a “historical relationship between accelerated investment periods and revenue reacceleration” at Amazon, noting the stock’s outperformance following such investment cycles.
Amazon’s shareholder letter emphasized the attractive free cash flow and return on invested capital profiles of data center investments. The company expects AI infrastructure costs to decrease as more chip options become available outside of Nvidia and foresee AI inference surpassing model training as the primary cost driver of AI.
Both Amazon and Alphabet are focusing on developing custom AI chips for inference. Amazon’s new Trainium2 chip is reported to have a 30% to 40% better price-to-performance ratio compared to current graphic processing units (GPUs). The company aims to reduce customer inference costs, believing this will boost overall AI spending.
Alphabet introduced its seventh-generation AI chip, Ironwood, designed explicitly for inference, with enhanced computational power and memory capacity. It marks Alphabet’s first chip tailored for inference and is the company’s most energy-efficient chip to date.
With AI infrastructure investment, both Amazon and Alphabet are positioning themselves for long-term success, particularly as Microsoft potentially slows its expenditure. The demand for cloud computing and AI services is rising as these companies assist customers in creating their own AI models and applications and running AI workloads on their platforms.
Both companies’ development of custom AI chips is instrumental in lowering AI infrastructure costs, with these chips consuming less power and designed specifically for inference tasks. AI is also becoming integral to their other business operations—Amazon uses AI to optimize its logistics and warehousing, while Alphabet has made advancements with its Gemini 2.5 model, enhancing its position in the AI landscape, aiding its search and advertising businesses with innovations such as the Veo 2 text-to-video generator.
Despite a recent market downturn, the stocks of both companies are trading at appealing valuations, and historical patterns indicate potential long-term success, marking them as promising long-term investments.
