This enhanced collaboration empowers enterprises with scalable AI infrastructure to innovate and optimize generative AI applications.
Google Cloud and NVIDIA have announced an enhanced their partnership to empowering the machine learning (ML) community with advanced technology to streamline the development, scaling and management of generative AI applications. This collaboration integrates Google Cloud’s global infrastructure with NVIDIA’s Grace Blackwell AI computing platform and DGX Cloud, alongside the NVIDIA Clara computing suite, to foster innovation across various sectors.
Empowering AI development with advanced computing platforms
Google is incorporating the NVIDIA Grace Blackwell AI computing platform DGX Cloud to further AI breakthroughs across its products and developer ecosystem. This move is part of a broader effort to provide developers with flexible tools and frameworks, including NVIDIA NIM inference microservices and support for JAX on NVIDIA GPUs.
Google Cloud CEO Thomas Kurian highlighted the partnership’s foundation in hardware and its extension through software and managed platforms. “Together with NVIDIA, our team is committed to providing a highly accessible, open and comprehensive AI platform for ML developers,” said Kurian.
NVIDIA CEO Jensen Huang stressed the importance of solutions that allow enterprises to leverage generative AI swiftly and efficiently. “With expanded infrastructure offerings and new integrations with NVIDIA’s full-stack AI, Google Cloud continues to provide customers with an open, flexible platform to easily scale generative AI applications,” Huang remarked.
Advancements in AI technology and Infrastructure
The collaboration brings several key advancements to the AI and ML community, including:
- NVIDIA Grace Blackwell adoption: Google Cloud will utilize the Grace Blackwell platform to enhance real-time inference capabilities for large language models (LLMs), offering these powerful instances to customers.
- Enhanced GPU support: Google Cloud’s integration with NVIDIA GPUs, including the H100 and L4 Tensor Core GPUs, supports the development and deployment of AI models through JAX and the Vertex AI instances, enhancing the training and inference of LLMs.
- Deployment of NVIDIA NIM: The integration of NVIDIA NIM inference microservices into Google Kubernetes Engine (GKE) aims to streamline generative AI deployment and ensure scalable AI inferencing across enterprises.
- Support for NVIDIA NeMo: Google Cloud’s adoption of NVIDIA NeMo across its platform facilitates the automation and scaling of generative AI model training and deployment, making it easier for developers to kickstart their AI projects.
Expanding support for generative AI applications
This partnership also extends support for AI-driven data science and analytics through Vertex AI and Dataflow, leveraging NVIDIA GPUs for scalable infrastructure and tooling. The AI Hypercomputer, powered by NVIDIA’s GPUs on Google Cloud, exemplifies a supercomputing architecture that enables AI researchers and developers to work with the most sophisticated AI models using optimized tools and frameworks.
Impact on industry innovators
The collaboration between Google Cloud and NVIDIA has already influenced industry innovators like Runway, Palo Alto Networks and Writer, enabling them to enhance model training and inference performance while optimizing costs and efficiency. These successes underscore the partnership’s potential to drive meaningful advancements in AI development and deployment.
Also read:
- Microsoft’s M12 Fund and GitHub Invest in Low-Code Platform ToolJet
- Microsoft-backed Builder.ai Secures Over US$250 Million in Series D Funding
- Google Docs vs. Microsoft Word: Which Is Better for Businesses?
Header Image from Freepik
Press release link: https://www.prnewswire.com/news-releases/google-cloud-and-nvidia-expand-partnership-to-scale-ai-development-302092022.html
