LiveRamp Holdings Inc. (NYSE: RAMP) announced it has integrated NVIDIA AI infrastructure into its clean room architecture to support compute-intensive AI workloads. The integration allows AI partners and brands to train and deploy models up to 15 times faster than CPU-based environments, according to the company.

The GPU-optimized infrastructure enables AI partners to use existing code within LiveRamp's clean rooms without requiring modifications for CPU-based systems. The technology supports model training and inference while maintaining data security and protecting proprietary model weights.

The system allows marketers to access advanced AI models without exposing first-party data outside the clean room environment. AI partners can run proprietary models to provide campaign optimization and audience selection services to brands while keeping raw data secure.

"This integration brings together world-class compute with the identity and data foundation needed to power the next generation of AI in advertising," said Pirouz Nilforoush, President and Co-Founder at inPowered AI.

Matt Karasick, Chief Product Officer at LiveRamp, stated that the upgrade makes high-performance computing accessible to the company's network of more than 900 brands, publishers, and platforms.

Jamie Allan, Director of AdTech & Digital Marketing Industries at NVIDIA, said the integration provides marketers with a foundation to scale marketing efforts and accelerate innovation.

The integration is currently in limited release, with general availability expected later this year. LiveRamp recently expanded its Marketplace to include data and models for AI training and access to AI-powered applications.

LiveRamp, headquartered in San Francisco, operates a data collaboration network connecting brands, retailers, financial services providers, and healthcare companies.