Amazon reports Q2 AWS revenue up 17.5% YoY to $30.9B, vs. $30.8B est., and AWS operating income up 8.8% YoY to $10.2B, vs. $10.9B est.
Margins fell 35.5% to 32.9% YoY despite charging profane rates for new things ($9 per resource per month that IAM Internal Access Analyzer watches?!) [embedded post] X: Daniel Newman / @danielnewmanuv : If $AMZN AWS growth rate of 17% stays consistent and $MSFT Azure 39% rate stays in place Microsoft will be the largest cloud in ~4 years. Austin Lyons / @theaustinlyons : Strong growth, but AWS is late to GenAI and still supply constrained. The risk is that Azure and GCP scale faster and lock in share before AWS can close the gap. From Andy Jassy on the AWS earnings call: “In the rapidly evolving world of generative AI, AWS continues to build Ben Bajarin / @benbajarin : Folks aren't going to like the AWS growth of 17% when Azure and GCP are both +30% YoY. The CSP landscape is defintely shifting driven by AI workloads. [image] Gene Munster / @munster_gene : I'm surprised $AMZN stock did not move higher on the comment that AWS was capacity constrained. Said backlog grew at 25%, suggesting that's the true growth rate.
Context & Ripple Effects
AWS has reaccelerated from the 12% growth reported in its Q2 2023 cloud results and the 13% reported in Q4 2023 AWS revenue, but this quarter’s growth remains below the 19% rate recorded in Q3 2024.
The key tension is not demand alone: reported revenue slightly cleared expectations while operating income and margin came in below them. Commentary in the coverage points to capacity constraints and a growing backlog, making infrastructure availability central to how quickly AWS can convert demand into revenue.
First-order effects
- AWS’s $30.9B in quarterly revenue confirms continued expansion, but the $10.2B operating-income result below estimates puts immediate focus on the cost of serving that growth and the margin decline.
- Capacity constraints can defer revenue recognition even where customer demand exists, leaving AWS to prioritize available infrastructure and customers to compete for scarce compute.
Second-order effects
- Azure and GCP have an opening to win workloads that need AI capacity sooner; sustained relative growth differences would matter more than AWS’s absolute revenue lead.
- Higher infrastructure costs and AWS’s expanded charges for newer services increase pressure on customers to govern AI and cloud usage, strengthening the case for Amazon’s broader profitability discipline to extend into cloud consumption economics.
Third-order effects
- Cloud competition is increasingly shaped by the speed of converting AI demand into deployed capacity, rather than by the legacy scale of a provider alone.
- If constrained supply and lower incremental margins persist, AI infrastructure could push cloud providers toward more utility-like investment cycles: large upfront build-outs, backlog-led planning, and more scrutiny of returns on capacity.
The trend: Hyperscale cloud competition is shifting toward an AI-capacity race in which supply availability and compute economics determine who can translate demand into durable share.