Datadog Unveils GPU Monitoring to Cut AI Costs

Datadog announces the launch of GPU monitoring within its observability stack, designed to provide AI-focused companies with deeper insights into their GPU-related expenditures, which currently constitute about 14% of cloud computing costs. This development comes on the heels of a report from IDC indicating that global AI infrastructure spending reached nearly $90 billion in late 2025, signifying a substantial growth of 62% over the previous year. Datadog's tool aims to facilitate faster troubleshooting and improve workload management by offering a comprehensive view of GPU fleet health, associated costs, and performance metrics.

The introduction of this monitoring tool marks a significant shift towards enhanced visibility for organizations investing heavily in AI technologies. As companies grapple with soaring GPU costs and seek to optimize their AI infrastructure, this capability could help companies not only in managing expenses more effectively but also in making informed decisions regarding their AI investments. This strategic enhancement positions Datadog as a critical player in the ongoing dialogue about AI infrastructure efficiency, ensuring that businesses can navigate the complexities of increasing AI operational costs with greater clarity and insight.