Nvidia is reportedly redirecting its attention from artificial intelligence (AI) chipsets to the field of robotics and embodied AI. The tech powerhouse, based in Santa Clara, is aiming to enhance processing capabilities and training for robots amid escalating competition in the AI hardware landscape. A recent statement from an Nvidia executive indicated the company’s intentions to begin training AI agents in simulated environments to better equip them for real-world applications. Earlier this year, Nvidia also introduced a new AI platform called Generalist Robot 00 Technology (GR00T) tailored for humanoid robotics.
Nvidia Looks Towards Robotics as an Untapped Space
A report by the Financial Times reveals that Nvidia is strategically positioning itself to become a leading platform for robotics processing and AI-driven simulation training. Deepu Talla, Nvidia’s Vice-President of Robotics, remarked to the publication that a transformative moment in physical AI and robotics is imminent.
This strategic pivot may lead to a reduced emphasis on AI chipsets and GPUs, which constitute the primary revenue stream for the company, valued at approximately $3.6 trillion (around Rs. 307.9 lakh crores). The report noted that this shift in focus was influenced by the intensified competition facing the AI semiconductor sector in recent months.
According to the report, Nvidia intends to establish itself as a comprehensive “full stack” solution for robotics, integrating both hardware acceleration and software necessary for training applications. Recently, Jim Fan, who leads Nvidia’s Embodied AI division, disclosed plans to train embodied AI agents—robots equipped with sensors and mechanical components—in simulated scenarios to prepare them for real-world challenges.
Nvidia is also on Technology News of launching Jetson Thor, a chipset developed through the company’s Isaac Robotics Platform, designed to power its GR00T platform for simulation-based training. This innovative platform utilizes generative AI capabilities to facilitate training for humanoid robots, significantly reducing the need for extensive coding.
Talla emphasized to the Financial Times that the decision to focus on the robotics sector stems from the advancements in generative AI models, allowing for robot training in simulated environments. He noted, “In the past 12 months, this gap has matured sufficiently that we can now carry out experiments in simulation, combining with generative AI, that we could not do two years ago. We provide the platform for enabling all of these companies to do any of those tasks.”
