LiveRamp Accelerates AI Clean Rooms with NVIDIA GPU Integration
Event summary
- LiveRamp has integrated NVIDIA’s AI infrastructure into its clean room architecture.
- The upgrade enables AI model training and inference speeds up to 15x faster.
- The integration allows AI partners to use existing code without rearchitecting for CPU-based environments.
- LiveRamp’s Marketplace now offers data and models for AI training, alongside AI-powered applications.
- General availability (GA) of the integration is expected later in 2026.
The big picture
LiveRamp's move to NVIDIA GPUs signifies a broader trend of AI model training shifting to specialized hardware, driven by the increasing computational demands of AI-driven marketing. This integration addresses a key bottleneck in the data collaboration workflow, allowing for faster iteration and more sophisticated AI applications within a privacy-preserving environment. The partnership positions LiveRamp to capitalize on the growing demand for AI-powered marketing solutions, but also increases its reliance on NVIDIA’s hardware ecosystem.
What we're watching
- Adoption Rate
- The speed of adoption among LiveRamp’s 900+ brands, publishers, and platforms will determine the immediate impact on revenue and market share.
- Competitive Response
- Competitors in the data collaboration and clean room space will likely accelerate their own GPU infrastructure investments to remain competitive.
- Model Security
- The long-term success hinges on LiveRamp’s ability to maintain robust data security and IP protection as the scale and complexity of AI models increase.
Related topics