AMD Announces Broad Hardware Support for Google's New Gemma 4 AI Models

AMD launched "Day One" support for Google’s Gemma 4 AI models. The models run on Radeon GPUs and Instinct datacenter GPUs. Support also includes Ryzen AI CPUs.

Developers can immediately perform inference tasks locally and in the cloud. AMD integrated the models with vLLM and LM Studio. The models also support the ROCm software platform.

These integrations facilitate broad accessibility and deployment. The move supports AMD's strategy to compete in the AI market via an open software ecosystem.