#bStocksCIS @BinanceCIS
NVIDIA's newest AI deal is not mainly about a new GPU. It is about reserving the bottlenecks that come before a GPU can even be installed: land, power and a completed data-center shell.
On August 17, NVIDIA said it will provide credit support for the first 4.25 IT-GW at SB Energy's PORTS-Pike campus in Ohio, with an option over the remaining 3.75 IT-GW. OpenAI is expected to use all 8 IT-GW. SB Energy will build, own and operate the site under a 20-year lease to OpenAI, while NVIDIA is investing $1.5 billion in SB Energy.
That changes how I read $NVDAB .
The positive interpretation is that NVIDIA is securing a long-duration route for its full AI stack into a very large customer deployment. Exclusivity strengthens that demand signal.
But these figures are not near-term chip revenue. Credit support is not a GPU purchase order. An option is not a firm commitment. Capacity is planned to arrive in phases beginning in 2028, so power delivery, construction, financing and customer concentration still sit between today's announcement and future compute sales.
The strategic shift matters most to me: NVIDIA is no longer waiting for data centers to become ready. It is helping make them financeable.
Does stronger demand visibility justify tying more capital and credit support to infrastructure years before it produces revenue?
NVIDIA's newest AI deal is not mainly about a new GPU. It is about reserving the bottlenecks that come before a GPU can even be installed: land, power and a completed data-center shell.
On August 17, NVIDIA said it will provide credit support for the first 4.25 IT-GW at SB Energy's PORTS-Pike campus in Ohio, with an option over the remaining 3.75 IT-GW. OpenAI is expected to use all 8 IT-GW. SB Energy will build, own and operate the site under a 20-year lease to OpenAI, while NVIDIA is investing $1.5 billion in SB Energy.
That changes how I read $NVDAB .
The positive interpretation is that NVIDIA is securing a long-duration route for its full AI stack into a very large customer deployment. Exclusivity strengthens that demand signal.
But these figures are not near-term chip revenue. Credit support is not a GPU purchase order. An option is not a firm commitment. Capacity is planned to arrive in phases beginning in 2028, so power delivery, construction, financing and customer concentration still sit between today's announcement and future compute sales.
The strategic shift matters most to me: NVIDIA is no longer waiting for data centers to become ready. It is helping make them financeable.
Does stronger demand visibility justify tying more capital and credit support to infrastructure years before it produces revenue?
