Morgan McGary’s Post

Introducing GR00T, a cutting-edge foundation model poised to revolutionize humanoid robot learning both in simulation and the real world. Powered by NVIDIA GPU-accelerated simulation, GR00T empowers humanoid embodiments to swiftly grasp new skills through imitation learning and reinforcement techniques using NVIDIA Isaac Lab. From just a few human demonstrations and video data, GR00T generates precise robot movements with ease. With GR00T, multimodal instructions and past interactions are seamlessly processed to produce actionable outputs for enhanced robotic performance. Get ready to witness the future of robotics unfold! #nvidia #robotics #ai

Nvidia CEO Jensen Huang captivates the audience at GTC conference by unveiling a squad of nine humanoid robots, marking the launch of "Project GR00T" - a ground-breaking initiative in robotics. In collaboration with renowned companies like Figure AI and Boston Dynamics, Nvidia's AI-driven robots are set to redefine everyday tasks through adaptive learning capabilities. Not stopping there, Huang introduces the cutting-edge Blackwell chip, promising enhanced AI and robotics performance. With Nvidia's relentless commitment to innovation, the future holds exciting advancements that will undoubtedly shape the landscape of AI and robotics. What are your thoughts on the potential impact of humanoid robots in our daily lives? #Nvidia #AI #Robotics #Innovation #ProjectGR00T #GTCConference

Robots will continue to expand and exhibit exponential growth behaviour too. If it seems like nothing is happening, then everything happens at once - that is a symptom of exponentials. AI will power this take off in capability: LLMs made human interactions with robots easier (asking and answering questions... driving actions based on vision generated prompts etc). Better architectures and world simulation models are then taking it to a whole new level. PS: did I mention new GPUs coming out too? https://lnkd.in/eXiSsqGR

Nvidia CEO Jensen Huang starts the GTC keynote in San Jose with numerous AI-infused chip announcements. Explore summary right here. Nvidia GTC 2024 Keynote 👉 https://lnkd.in/ecQmDF7N #AI #GTC2024 #tech #robotics #machinelearning #artificialintelligence #datascience

I just finished watching NVIDIA's GTC keynote, and it was a glimpse into the future of accelerated computing. While robots weren't the main focus, their advancements in AI, from training to deployment, are sure to revolutionize automation across industries. This technology has the potential to empower next-gen robots for complex tasks. I'm excited to see how NVIDIA's innovations will shape the future of robotics! #NVIDIA #GTC #AI #Robotics https://lnkd.in/dBuYibNs

NVIDIA Omniverse NVIDIA created virtual worlds simulations where robots learn to make decisions by performing actions in a virtual world that obey the laws of physics. #generativeai #Omniverse #NVIDIA #computex2024 #robotics #machinelearning #outsourcing

“The era of #robotics has arrived. Everything that moves will one day be autonomous,” said Jensen Huang, founder and CEO of NVIDIA. “We are working to accelerate generative physical #AI by advancing the #NVIDIA robotics stack, including #Omniverse for simulation applications, #ProjectGR00T humanoid foundation models and the #JetsonThor robotics computer.” The #Isaac platform features a suite of NVIDIA-accelerated libraries, #AI foundation models and simulation technologies that are available today to #robot makers to integrate into their technology stacks. #Robotics #AI #Omniverse #NVidia #NVidiaIsaac #GenerativePhysicalAI #Computacenter https://lnkd.in/edPpPhQm

From my writeup titled “ AI is seeking the edge “ from 2017 , https://lnkd.in/fm52KDz “Race is on to produce low-power, deep learning hardware accelerators on embedded systems. Edge devices and environments would be equipped with custom AI chips. Computer vision is already pluggable. I still see the relevance of the traditional client-server model remain as is except that "client" and "server" will trade places. "Edge" is the new server and the cloud services, the new clients. They might occasionally talk:)” Jetson is emerging as the edge first AI platform . And edge is where the resistance to centralization would begin 💪 ...For those living on the edge , never been a better time to build truly decentralized edge and near edge AI powered apps.

As we approach the end of the year, it’s time to look back at the most popular stories from the NVIDIA Technical Blog in 2023. Read the top posts on groundbreaking research and developments in #generativeAI, #LLM, #HPC, #robotics, and virtual worlds.

🚀 Exciting Announcements from NVIDIA GTC Keynote by Jensen Huang!      🧠 Blackwell Computing Platform: Revolutionizing real-time generative AI and large language models for the trillion parameter LLM era.    📦 NIM-Microservices: Streamlining AI deployment with cutting-edge microservices.    ☁️ Omniverse: Unveiling new simulation capabilities and also CloudXR streaming for the Apple Vision Pro.    🌍 Earth 2: Model for weather simulation with high resolution also leveraging Gen AI.    🤖 Project GROOT: Introducing groundbreaking multimodal foundation model not only for humanoid robotics.   What are your highlights of #GTC2024 so far?