“A rising tide lifts all boats,” goes the popular aphorism. That is certainly true when it comes to the AI-driven results of the three major U.S. hyperscalers this quarter. Alphabet, Amazon, and Microsoft all managed to post double-digit growth figures, with cloud revenue as the primary driver. Hardware players are also benefiting, while earlier missteps by AI players like Meta are still being punished.
The most striking result came from Alphabet, better known as Google’s parent company. The company was once reliant on revenue from its ubiquitous search engine, but has since established itself as a serious cloud player with substantial revenue. 18 percent of Alphabet’s total revenue comes from Google Cloud, which posted a staggering growth rate of 63 percent compared to the same quarter last year.
This astronomical growth would not have been possible without $35.67 billion in capital expenditures. These capex costs, largely driven by the expansion of AI infrastructure, are estimated to reach $175 to $185 billion by the end of the year. Fortunately for Alphabet, the costs for AI outputs are dropping significantly: Gemini 3 has led to a 30 percent cost reduction. The fact that the Gemini app now switches the “Pro” setting to a default “Fast” option even for paying customers likely has something to do with that as well.
AWS and Microsoft are following suit
AWS remains the public cloud leader, with quarterly revenue of $181.5 billion, $37.6 billion of which comes directly from the cloud. At 28 percent, AWS’s growth is significantly lower than that of Google Cloud, but that’s only logical given that AWS holds the largest market share. It’s striking that cloud revenue was actually highest at Microsoft this quarter: $54.5 billion. Azure grew by 40 percent compared to the same quarter in 2025. However, the revenues from “Microsoft Cloud” are bundled with Office 365, Dynamics 365, and LinkedIn. If you factor that in, you can safely assume that AWS still generates the highest revenue
Like Alphabet, AWS and Microsoft are opting for massive investments in AI infrastructure. We already knew that, but the figures are new: $44.2 billion from AWS, a 76 percent increase compared to Q1 2025, and Microsoft spent $31.9 billion, 49 percent more than the same period a year ago. Next quarter, Microsoft expects to spend $40 billion on capex, with $190 billion for all of 2026.
Hardware benefits
Capex spending is directed toward various parties. The main cost item is hardware. It is no coincidence that Samsung is reporting impressive figures. According to Samsung CEO Kim Jaejune, the gap between supply and demand for AI memory will only widen in 2027. Together with Micron and SK Hynix, which reported revenue growth of 198 percent (!), Samsung is reaping the benefits of this shortage. Samsung’s margin is at a record high of 72 percent.
Chipmaker Qualcomm is also on the rise, largely thanks to hyperscaler demand. Data center chips for “a major hyperscaler” will be delivered this calendar year, an intriguing and significant sign that Arm-based AI chips can take over the workloads of GPUs. This will likely be only partial. It is striking that Google Cloud is beginning to emerge as a competitor to Qualcomm, as that cloud division will supply some customers with its own TPUs. A comprehensive analysis of the latest generation of these can be found below:
Read: Google unveils TPU 8t and TPU 8i chips; splits training and inference
Not everyone is a winner
There must be a difference, and Meta and IBM are recent examples of tech companies that had to swallow a stock market correction after presenting their quarterly results. Despite ongoing revenue from social media platforms and AI plans, Meta still has to deal with the metaverse debacle, in addition to the fact that capital expenditures for AI are a cause for concern. Unlike hyperscalers, Meta only generates revenue from this infrastructure by building models or using the hardware for its own AI workloads. This profile, combined with the billions lost on the metaverse venture through Reality Labs, led to a 6 percent stock drop.
IBM is in a similar situation, where positive figures from the mainframe business, among others, contrasted with declines in sectors that typically have higher margins. It remains to be seen whether the company can leverage AI through consulting using its own technologies, among other things. That consulting division grew by only 1 percent. The 9.5 percent in total revenue growth stands in sharp contrast to the 9 percent drop in market value that followed the presentation of the quarterly results. It shows that the AI expansion is by no means rewarding everyone, especially if the hardware itself is not directly focused on AI.