Nvidia AI boom faces slowdown amid financing push

Nvidia’s Q2 2026 revenue hit a record $81.6 billion, lifting the stock about 12 % YTD, while analysts warn that custom silicon from hyperscalers could curb future GPU demand. AMD’s new AI‑chip family prompted higher price targets, and Intel’s data‑center AI division surged 59 % YoY, underscoring a fiercely competitive market.

The company is negotiating AI‑infrastructure deals exceeding $750 billion, including roughly $500 billion linked to SK Hynix. New details show Nvidia is also financing customers, co‑building data‑center projects and providing guarantees that could total $250 billion for OpenAI’s U.S. compute‑lease and $350 billion for chip purchases. This “circular economy” model ties Nvidia’s own financing activities to the hardware it supplies, raising concerns that demand could be artificially inflated and that the financial health of AI firms may become dependent on Nvidia’s credit extensions. Analysts note comparable financing structures, such as Google’s $35 billion guarantees for Anthropic, could introduce systemic risk if AI‑related capital spending slows.

Meanwhile, CUDA’s share of frontier‑lab model training has fallen about 70 % over two years as Google TPUs, Amazon Trainium and other custom silicon gain traction, though Nvidia’s high‑speed networking (NVLink, InfiniBand) remains a durable advantage. The broader AI boom continues to lift related technology stocks and even crypto tokens, but risks from capital‑intensive spending and emerging financing models keep the outlook uncertain.