Date: August 26, 2026 Main source: NVIDIA Link: https://nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-second-quarter-fiscal-2027 Supplementary sources: https://nvidianews.nvidia.com/news/aws-and-nvidia-to-deliver-2-million-additional-gpus-and-next-generation-infrastructure-for-agentic-and-physical-ai
Summary
Nvidia reported quarterly revenue of $96.2 billion, up 106% year over year, with $89 billion coming from data centers. The company projects about $108 billion in revenue for the next quarter. The same day, AWS and Nvidia announced they will deploy 2 million additional Nvidia GPUs across Amazon's global infrastructure, and expand their collaboration on Vera CPUs, networking, Nemotron models and robotics. The move confirms that demand for infrastructure for training, inference, agents and physical systems keeps expanding.
Why it matters
The business thesis behind massive AI spending is moving from experimentation to productive capacity. The large cloud providers are securing infrastructure several years ahead because they expect inference and agents to generate recurring workloads. For the rest of the market, this also means that the cost, availability and efficiency of compute will remain important competitive factors.
Practical implications
- Design applications so you can switch between models and infrastructure providers without redoing the whole architecture.
- Measure cost per completed task, not just cost per token.
- For small companies, prioritize APIs and managed services over buying your own infrastructure.