Amazon is raising prices for GPU rental capacity on AWS as of July 1, marking the second price increase in six months and reflecting continued tight supply for AI computing resources. The move tests whether customers will pay more for cloud-based AI compute or seek cheaper alternatives, with implications for AWS margins and the broader cloud market's ability to monetize AI infrastructure.
Summaries like this, in your inbox every morning.
Sign up free →What happened
Amazon Web Services announced new pricing for EC2 Capacity Blocks for machine learning, effective July 1, with hourly rates per accelerator ranging from $5.191 for P5 in U.S. regions to $14.04 for P6-B300. The company said reservation prices are updated periodically based on supply and demand, with the last increase occurring in January.
Why it matters
The price increases suggest that demand for AI computing power remains tight across key cloud regions, signaling that Nvidia-based GPU compute is still a scarce resource. Higher AWS GPU pricing could support cloud revenue and margins if demand holds, but customers may respond by shifting workloads to cheaper internal chips and alternatives.
What to watch
The next key question is whether customers absorb the increases or move more workloads to competing solutions—a measure of how much pricing power AWS actually has in the competitive AI infrastructure market.
No discussion yet for this article
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
1 minute a day. The AI essentials.
200+ sources · Email / LINE / Slack