A $30B valuation wiped in under 48 hours. Nvidia-backed Australian AI data center firm Firmus Grid pulled its IPO after investors balked at the price.
The company wanted to raise ~$5.5B at nearly 3X its valuation from just three months prior. The market said no.
Why? According to the FT, Firmus had only 46 MW of operational capacity out of over 900 MW contracted. Investors were being asked to pay a massive premium for infrastructure that barely exists yet.
This feels like a pricing disaster, not a referendum on AI infrastructure demand. The sector remains red-hot, but slapping a $30B price tag on ambitious expansion plans without meaningful execution is delusional.
There's a massive gap between contracted power, actual deployments, and real revenue. That's why I focus on execution, customer traction, financing strength, and operational capacity when evaluating names like $NBIS, $CRWV, and $IREN.
The AI infrastructure boom can keep running while individual companies crater on valuation disconnects. Both are true. Price discipline matters, even in a bull market.
The company wanted to raise ~$5.5B at nearly 3X its valuation from just three months prior. The market said no.
Why? According to the FT, Firmus had only 46 MW of operational capacity out of over 900 MW contracted. Investors were being asked to pay a massive premium for infrastructure that barely exists yet.
This feels like a pricing disaster, not a referendum on AI infrastructure demand. The sector remains red-hot, but slapping a $30B price tag on ambitious expansion plans without meaningful execution is delusional.
There's a massive gap between contracted power, actual deployments, and real revenue. That's why I focus on execution, customer traction, financing strength, and operational capacity when evaluating names like $NBIS, $CRWV, and $IREN.
The AI infrastructure boom can keep running while individual companies crater on valuation disconnects. Both are true. Price discipline matters, even in a bull market.