Google and NVIDIA Expand AI Partnership to Tackle Real-World Challenges

How Google and NVIDIA are teaming up to solve real-world problems with AI

From Google DeepMind's AI advancements, to Google Cloud's AI-first infrastructure, to our Other Bets' use of AI to tackle complex challenges in areas like healthcare, energy and robotics -- every corner of our company is focused on pushing the frontier of AI and bringing its benefits to everyone. NVIDIA is a critical partner in this mission. As a world leader in AI and accelerated computing, NVIDIA has collaborated with Google and Alphabet for years on a wide variety of projects, including Android, advanced AI research, next-generation hardware and software optimizations, and beyond -- all with the goal of making AI more accessible for developers. This week at the NVIDIA GTC global AI conference, we're doubling down on our partnership with NVIDIA with announcements across Google and Alphabet: Training large AI models requires serious computing power, and NVIDIA continues to raise the bar on GPU performance. We're excited to announce our A4 VM is now generally available, based on NVIDIA HGX B200 GPU -- and our A4X VMs based on NVIDIA GB200 NVL72 will be available soon, too. We're committed to supporting the latest Blackwell GPUs, including the just-announced NVIDIA RTX PRO 6000 Blackwell and NVIDIA GB300. This means faster training times, smoother deployments and the ability to tackle even more complex AI challenges. Google DeepMind's Gemini is the result of decades of large language model research and is our most capable AI model. This week we'll share how rapid advances in Gemini are designed to support more developers and end users -- and how our collaboration with NVIDIA is helping to make that possible. The Google DeepMind team also released Gemma 3 last week, the latest in our collection of lightweight, state-of-the-art open models. NVIDIA played a key role in the development of these open models, working closely with our researchers to optimize Gemma to run on its vast accelerated computing ecosystem. This means developers can easily access the same technology that powers Gemini and implement it on any NVIDIA GPU available to them. Separately, NVIDIA also announced today that it has employed Google DeepMind's SynthID watermarking tool on its Cosmos video generation platform. This marks our first external deployment of the technology, and we're proud to help instill user trust in AI-generated content. Our collaboration with NVIDIA isn't just about building AI tools to help in daily life; it's about fundamentally improving the quality of life for everyone. That means working together to apply AI to address some of the world's most complex challenges. For starters, we're teaming up to apply the most advanced frontier models to areas from energy to drug discovery. We've always believed that the value of technology lives in its capacity to benefit people everywhere. That's been foundational to our mission and it's what our growing collaboration with NVIDIA is built on. 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 AI more accessible and helpful to developers and users, and collaborating in ways that will drive innovation for years to come. If you're attending GTC this week, swing by Google Cloud's booth #914 and check out one of the many Google sessions to learn more about our work together.

Google, Nvidia partner to apply AI in robotics, drug discovery and energy grids

Google Cloud will also be among the first to adopt Nvidia's GB300 NVL72 and RTX PRO 6000 Blackwell Server Edition GPUs. Additionally, Nvidia will adopt SynthID, a digital watermarking tool from Google DeepMind that embeds invisible markers in AI-generated images, audio, text and video. Nvidia will be the first industry partner to implement the tool. Google is working with graphics processing unit (GPU) maker Nvidia to apply artificial intelligence (AI) and simulation technologies to address real-world use cases such as robotics, drug discovery and energy grid optimisation. Energy grids: Tapestry, Google X's electric grid project, and Nvidia are developing faster and more accurate electric grid simulations. Drug discovery: Isomorphic Labs, a subsidiary of Google parent Alphabet, and Nvidia are using AI to speed up the discovery of new medicines. Robotics: Intrinsic, also part of the Alphabet family, and Nvidia are working on improving robots' ability to grasp objects. Simulation tools: Google DeepMind and Nvidia have introduced MuJoCo-Warp, an open-source physics simulator aimed at supporting robotics research. Google Cloud will also be among the first to adopt Nvidia's GB300 NVL72 and RTX PRO 6000 Blackwell Server Edition GPUs. Additionally, Nvidia will adopt SynthID, a digital watermarking tool from Google DeepMind that embeds invisible markers in AI-generated images, audio, text and video. Nvidia will be the first industry partner to implement the tool. Alphabet CEO Sundar Pichai said on X, "Our ongoing and deep partnership with @NVIDIA spans the early days of Android to today's cutting-edge AI collaborations across Alphabet. I'm really excited about the next phase of our partnership as we work together on agentic AI, robotics, and bringing the benefits of AI to more people around the world. #GTC25." "Alphabet and NVIDIA have a longstanding partnership that extends from building AI infrastructure and software to advancing the use of AI in the largest industries," said Jensen Huang, founder and CEO of Nvidia. "It's a great joy to see Google and NVIDIA researchers and engineers collaborate to solve incredible challenges, from drug discovery to robotics."