$NVIDIA(NVDA)$ While Jensen Huang’s comments on AGI may invite discussion, they reflect the pace at which AI capabilities are advancing. Systems are more flexible and broadly useful than ever, and NVIDIA’s infrastructure has been a key driver behind that progress.
$NVIDIA(NVDA)$ This partnership involves AES, Constellation, Invenergy, NextEra Energy and Vistra, focusing on constructing a novel AI factory that connects directly to the power grid. These facilities will operate using Nvidia's Vera Rubin DSX design and DSX Flex software, enabling quicker integration into grid services while also serving as flexible energy assets. In simple terms, these are not just data centres; they are designed to both consume and support electricity systems.
Competition is certainly increasing with $Advanced Micro Devices(AMD)$ and custom chips from major tech firms such as $Meta Platforms, Inc.(META)$ and $Alphabet(GOOGL)$ , yet it still seems secondary. The real advantage is not merely hardware; it is $NVIDIA(NVDA)$ 's ecosystem. That is what maintains their lead, and while others are catching up on chips, they remain behind on the platform that actually runs AI at scale.
For those who still do not understand, even when $NVIDIA(NVDA)$ is not green for a minute, the vast companies they invest in and use their product are increasing in value, which we longterm holders enjoy on every quarterly earnings report.
$NVIDIA(NVDA)$ I observe that $IREN Ltd(IREN)$ has announced agreements to purchase over 50,000 B300 GPUs from Nvidia, a move expected to expand its total GPU fleet to around 150,000 units.
Due to this brief "war", $NVIDIA(NVDA)$ is likely to sell millions more chips. $NVIDIA(NVDA)$ stands to make billions of dollars from this "war" as their chips will be destroyed with each launch.
What makes $NVIDIA(NVDA)$ hard to replace isn’t just GPU performance. It’s the integration with memory, networking, CUDA tooling, and hyperscaler architecture. That ecosystem lock-in doesn’t disappear in the short term.