Nvidia’s $100B AI Gamble Sends Chips Soaring Even as Payoff Stays Elusive

Silicon Valley’s latest love affair just got complicated. Nvidia’sNVDA

$100B commitment to OpenAI ignited global chip stocks, sending Taiwan SemiconductorTSM

, SK Hynix, and some European peers soaring as investors bet on a tidal wave of AI infrastructure demand. Yet beneath the market euphoria lies a troubling disconnect — while companies can’t stop talking about AI, most struggle to articulate exactly how it’s boosting their bottom lines.

• Nvidia’s deal with OpenAI requires systems that demand 10 gigawatts of power — equivalent to roughly 4M to 5M graphics processing units (GPUs), according to CEO Jensen Huang.
• Meeting demands on that scale won’t come cheap, with Bain & Company projecting AI firms will fall $800B short of the $2T revenue needed by 2030.

AI’s shaky footing: FT’s review of S&P 500 filings reveals that while 87% of earnings calls painted AI in glowing terms, the same firms emphasized risks in their regulatory documents. Gartner’s Haritha Khandabattu argues much of the AI adoption rush stems less from long-term strategy than corporate “fear of missing out.” With tech giants set to pour over $500B a year into AI infrastructure by the early 2030s, the chip sector’s current celebration may prove premature if the AI revenue engine can’t catch up to its computational appetite.