NVIDIA and Google Cloud Empower AI Builders with New Tools and Resources

The collaboration between NVIDIA and Google Cloud is unlocking new frontiers for artificial intelligence developers. Through their joint community, both tech giants offer a robust ecosystem of learning, hands-on labs, and events designed to empower the creation of cutting-edge AI applications using NVIDIA's full-stack AI platform on Google Cloud.

Launched at Google I/O last year, this community has become an essential meeting point for developers, data scientists, and machine learning engineers. Its objective is clear: to enable professionals to sharpen their AI skills with the latest technologies from NVIDIA and Google Cloud, thereby driving innovation in the sector.

This year, the community is adding new features that promise to further enrich the member experience. These include a new learning path focused on using the JAX library on NVIDIA GPUs, an NVIDIA Dynamo codelab centered on inference optimization, and the continuation of their monthly developer livestreams.

Over the past year, the community has proven to be a go-to hub for AI builders leveraging NVIDIA-accelerated tools. This has led to the development of production-ready 'retrieval-augmented generation' applications on Google Kubernetes Engine (GKE), as well as the instrumentation of observability for agent workloads.

Developers within this community are also actively exploring the latest large language model (LLM) research. They are prototyping hybrid inference solutions, combining on-premises and cloud environments, for real-world use cases such as sports analytics and enterprise data pipeline creation.

NVIDIA and Google Cloud are equipping developers with learning resources and hands-on labs that integrate NVIDIA's libraries, open models, and tools with Google Cloud's AI platform. This enables the agile construction of optimized, production-ready AI applications.

For instance, developers can accelerate their data science and analytics workflows using the NVIDIA cuDF library in Google Colab Enterprise or Dataproc. They can also deploy multi-agent applications by combining Google DeepMind's Gemma models, NVIDIA Nemotron open models, and the Google Agent Development Kit, all running on Google Cloud's G4 VMs powered by NVIDIA RTX PRO 6000 Blackwell GPUs.

The collaboration extends to open frameworks like JAX, allowing developers to build, scale, and productize JAX workloads on NVIDIA's AI infrastructure on Google Cloud. This spans from single-GPU experiments to multi-rack deployments, ensuring robust performance and a consistent experience.

In the realm of responsible AI, NVIDIA and Google Cloud are collaborating closely. NVIDIA has become the first industry partner to work with Google DeepMind on SynthID, a digital watermarking technology for AI-generated content. This helps preserve the integrity of outputs from NVIDIA Cosmos world foundation models, which offer advanced 3D perception and simulation capabilities for physical AI systems.

The alliance between NVIDIA and Google Cloud is focused on ensuring AI builders have the infrastructure, software, and learning resources needed to make the most of the new agentic experiences and tools that Google I/O is highlighting this year. The skills and tools developers acquire in this community can scale effortlessly, taking projects from prototype to enterprise-grade workloads.