What Are the Right Sustainability Metrics for AI?

As AI accelerates datacenter demand, traditional metrics are no longer enough. SDEA president Babak Falsafi joined industry leaders in calling for transparent measurement frameworks that reflect the true impact of AI infrastructure.

AI and HPC workloads are reshaping how datacenters are built, powered, and cooled – yet the industry still relies heavily on outdated metrics like PUE. In a recent DatacenterDynamics broadcast, panelists from K2 Datacentres, Clear Sustainability, and SDEA highlighted the critical gaps in today’s sustainability reporting. From water usage to carbon intensity, from server utilization to energy provenance, the discussion made clear that current frameworks often ignore what matters most.

Babak Falsafi emphasized the need for independently verified, end-to-end metrics that address both efficiency and emissions across infrastructure and IT workloads. The SDEA Navigator is designed precisely for this purpose – going beyond point-in-time benchmarks to deliver a comprehensive, auditable view of sustainability.

As regulatory pressure increases and ESG scrutiny intensifies, only measured and verified data will separate real leadership from vague ambition.

Watch on DCD