Why Performance per Watt Is the Ultimate Metric for AI Infrastructure Efficiency
NVIDIA Shruti Koparkar
NVIDIA argues that performance per watt is the critical efficiency metric for AI infrastructure, as it determines how many tokens can be generated within a fixed power budget. The NVIDIA Blackwell NVL72 platform delivers up to 25x better performance per watt compared to the Hopper generation when serving mixture-of-experts models across a 72-GPU domain. This efficiency advantage will determine which organizations can scale their AI operations in a power-constrained environment.
Why it matters
Power is AI infrastructure’s inescapable constraint. How many tokens an AI factory can generate within a fixed power budget determines its revenue and profitability. Because of this, performance per watt — a metric that can’t be gamed, only earned through real-world results — is the foundation for AI factories. As agentic AI drives token demand […]
