Nvidia CEO Jensen Huang recently announced a strategic collaboration with Google to transform cloud infrastructure for the era of generative AI. The two tech giants aim to leverage their deep expertise in computer science and computing to reinvent cloud infrastructure and enable the widespread adoption of generative AI technologies.
Generative AI has the potential to revolutionize every layer of the computing stack, and this partnership seeks to address the unique challenges posed by this emerging field. By starting at every layer, from the chips to the software, Nvidia and Google intend to accelerate the development and deployment of AI models on the Google Cloud Platform’s Vertex AI.
To kick off their collaboration, the companies unveiled several new offerings at the Google Cloud Next conference. These include integrating Google’s serverless Spark offering with Nvidia GPUs, introducing new Google A3 VM instances powered by Nvidia H100 Tensor Core GPUs, and launching PaxML—Google’s framework for building large language models optimized for Nvidia accelerated computing.
PaxML represents a significant milestone in their joint efforts and provides developers with advanced experimentation capabilities and scalability for large language models. By reengineering and re-optimizing the software stack, Nvidia and Google are pushing the boundaries of AI research and enabling massive next-generation models while improving efficiency in terms of time, cost, and energy.
In addition, Google Cloud will be one of the first entities worldwide to access the Nvidia DGX GH200 AI supercomputer, powered by the Nvidia Grace Hopper Superchip. This cutting-edge hardware will empower Google and Nvidia to explore the potential of generative AI workloads.
The collaboration doesn’t stop there. Nvidia’s DGX Cloud AI supercomputing and software will be made directly accessible to Google Cloud customers, ensuring high performance and scalability for advanced training workloads. This integration will provide instant benefits to users of Google Cloud Platform and foster further innovation in the AI space.
Google Cloud CEO Thomas Kurian expressed his enthusiasm for the partnership, emphasizing the strategic value it brings to both companies. As AI technology evolves, the hardware and software infrastructure must adapt accordingly, making this collaboration instrumental in offering customers a wealth of accelerator choices.
The Nvidia-Google alliance represents a significant milestone in the field of generative AI and sets the stage for the continued innovation, integration, and development of AI technologies in the cloud. By combining their respective strengths, Nvidia and Google are poised to shape the future of AI infrastructure and propel the industry forward.
1. What is generative AI?
Generative AI refers to the use of artificial intelligence models to generate new and unique content, such as images, videos, music, and text, rather than simply recognizing patterns in existing data.
2. How will Nvidia and Google collaborate?
Nvidia and Google will collaborate closely to reinvent cloud infrastructure for generative AI. They plan to work on various aspects, starting from the chips and hardware to the software stack, with the goal of accelerating the development and adoption of AI models on the Google Cloud Platform.
3. What is PaxML?
PaxML is a new framework developed by Google and Nvidia for building large language models. It enables developers to leverage Nvidia’s H100 and A100 Tensor Core GPUs for advanced experimentation and scalability.
4. What is the Nvidia DGX GH200 AI supercomputer?
The Nvidia DGX GH200 AI supercomputer is a revolutionary computing system powered by the Nvidia Grace Hopper Superchip. It offers high-performance computing capabilities specifically designed for generative AI workloads.
5. How will customers benefit from this collaboration?
Customers of Google Cloud Platform will benefit from improved access to advanced AI infrastructure and accelerated training workloads. The collaboration between Nvidia and Google aims to drive innovation, efficiency, and scalability in the field of AI, ultimately benefiting developers, researchers, and end-users.