Google, NVIDIA Join Forces to Tackle Real-World Issues with AI

From Google DeepMind's strides in AI to Google Cloud's AI-centric infrastructure, and our Other Bets harnessing AI for challenges in healthcare, energy, and robotics—every part of our company is dedicated to advancing AI and making its benefits accessible to all.

NVIDIA plays a crucial role in this endeavor. As a global leader in AI and accelerated computing, NVIDIA has been collaborating with Google and Alphabet for years on numerous projects, including Android, cutting-edge AI research, next-gen hardware and software optimizations, and more—all aimed at making AI more accessible to developers.

This week at the NVIDIA GTC global AI conference, we're ramping up our partnership with NVIDIA with exciting announcements across Google and Alphabet:

Google Cloud’s AI infrastructure will help transform businesses with NVIDIA’s latest GPUs

Training large AI models demands hefty computing power, and NVIDIA keeps pushing the limits with their GPU performance. We're thrilled to announce that our A4 VM, powered by the NVIDIA HGX B200 GPU, is now generally available, and our A4X VMs, based on the NVIDIA GB200 NVL72, will be available soon. We're fully committed to supporting the latest Blackwell GPUs, including the newly announced NVIDIA RTX PRO 6000 Blackwell and NVIDIA GB300. This means quicker training times, smoother deployments, and the ability to tackle even more complex AI challenges.

Google DeepMind and NVIDIA will make the power of Gemini available to more people

Google DeepMind's Gemini is the culmination of years of large language model research and is our most advanced AI model yet. This week, we'll showcase how Gemini's rapid advancements are designed to support more developers and end users—and how our collaboration with NVIDIA is making this possible.

Last week, the Google DeepMind team also released Gemma 3, the latest in our series of lightweight, state-of-the-art open models. NVIDIA was instrumental in developing these open models, working closely with our researchers to optimize Gemma for their extensive accelerated computing ecosystem. This means developers can easily tap into the same technology that powers Gemini and use it on any NVIDIA GPU at their disposal.

Separately, NVIDIA announced today that it's using Google DeepMind's SynthID watermarking tool on its Cosmos video generation platform. This is our first external deployment of the technology, and we're proud to help build trust in AI-generated content.

Alphabet and NVIDIA will use AI to tackle the world’s most complex issues

Our collaboration with NVIDIA isn't just about creating AI tools for everyday use; it's about fundamentally enhancing the quality of life for everyone. This means working together to apply AI to some of the world's most pressing challenges. Here's what we're starting with:

• Smarter energy grids. Tapestry, X's moonshot for the electric grid, and NVIDIA are exploring ways to boost the speed and accuracy of electric grid simulations.
• Improved drug discovery. Isomorphic Labs and NVIDIA are pushing forward the development of new medicines using AI.
• More capable robots. Intrinsic and NVIDIA are making robots smarter and more capable, integrating NVIDIA's Isaac foundation models for better grasp capabilities.
• Advanced robotics simulation. Google DeepMind and NVIDIA are launching MuJoCo-Warp, a new open-source physics simulator to speed up robotics research.

We've always believed that technology's true value lies in its ability to benefit people everywhere. This belief is at the core of our mission and drives our growing collaboration with NVIDIA. By combining Google's expertise in AI research and infrastructure with NVIDIA's leadership in accelerated computing, we're committed to strengthening the foundation of AI, making it more accessible and useful to developers and users, and fostering innovation for years to come.

If you're at GTC this week, stop by Google Cloud's booth #914 and catch one of the many Google sessions to learn more about our joint efforts.