Overview
- Nvidia reported a record fiscal second-quarter with $96.2 billion in revenue and $89 billion from data-center sales, and CEO Jensen Huang said the company targets roughly 70% revenue growth next year.
- The company is packaging GPUs, its Grace and Vera CPU families, networking and partner financing into full data-center platforms to lock in customers and capture higher system revenue.
- Rising costs and bottlenecks in high-bandwidth memory, advanced packaging and infrastructure are squeezing margins and slowing how fast customers can deploy new AI clusters.
- Major customers remain large Nvidia buyers but are publicly hedging: OpenAI said it will diversify suppliers while still holding multiyear Nvidia compute commitments, and hyperscale cloud providers are developing custom accelerators.
- Nvidia is extending its influence with multibillion-dollar investments in partners and strong OEM demand, exemplified by Dell’s $95 billion AI server backlog, even as AI-safety warnings and a DOJ probe add regulatory and reputational uncertainty.