英伟达为 OpenAI 俄亥俄数据中心兜底 1050 亿美元

This $105B cap shows NVIDIA's pragmatism in scaling AI while managing risk, not a retreat. The push for massive GPU deployment looks like the real story here, not the deal size.

Ohio build starts with a $105B guarantee structure; NVIDIA's AI infrastructure plays a massive role in scaling GPU deployments down the road.

NVIDIA's $105B cap structure and SB Energy investment directly support the 4.25 GW Ohio build. If utilization hits expected levels, that 1.5M GPU deployment could logically drive $150B+ revenue for the next cycle.

Data center starts in Ohio, but NVIDIA's $105B cap limits downside risk while still allowing huge upside if they deploy 1.5M GPUs.

The $105B liability cap is a pragmatic risk control measure, signaling NVIDIA's confidence in OpenAI's project viability despite the downgrade from $250B. With the 4.25 GW facility now breaking ground, the focus shifts to execution. If deployment scales as Jensen suggested—using 1.5M GPUs per generation—the $150B to $200B revenue potential validates this structured approach. The move essentially turns a theoretical risk into a managed operational gap, paving the way for SB Energy's $1.5B investment to materialize.

NVIDIA's $105B cap looks tight vs. that $150B-$200B revenue target; I'll adjust my size if they breach the guarantee.