Nvidia stock plunged alongside AMD and Broadcom, part of a broader 5% chip sector drop after reports OpenAI may postpone its IPO to 2027. Amazon is raising AWS EC2 ML GPU rates from $5.19 to $14.04 per hour for P5 and P6 Nvidia instances starting July 1, while industry groups are exploring futures contracts to benchmark heterogeneous AI compute assets.
Nvidia shares fell sharply this week alongside AMD, Broadcom and other AI-linked names as reports surfaced that OpenAI may delay its planned public offering until 2027. The news triggered a roughly 5% decline across semiconductor and infrastructure stocks, reflecting investor caution toward AI spending and memory-chip demand.
Amazon will implement new EC2 Capacity Block rates for machine learning on July 1, with hourly prices per Nvidia accelerator rising to $14.04 for P6-B300, $12.355 for P6-B200, $5.191 for P5, $5.97 for P5e and $6.865 for P5en in U.S. regions. This marks the first GPU rental price increase since January, underscoring tight supply and strong demand for Nvidia-based compute resources.
Industry groups and startups are evaluating the creation of a financial market for AI compute, but face significant hurdles due to the heterogeneity of GPU offerings. Even identical Nvidia H100 chips can vary two- to fourfold in rental cost based on data center infrastructure, power quality and service stack, complicating efforts to establish a clean, benchmarkable futures contract.
Finance
