NVIDIA reportedly shifted its focus from artificial intelligence (AI) chipsets to robotics and the AI embodied. The Santa Clara-based tech giant is said to be seeking to provide machining capabilities and training for robots, as competition in the field of AI hardware continues to rise. It is worth noting that an NVIDIA executive recently highlighted the company’s plans to start training AI agents in preparation for the real world. Additionally, earlier this year, the tech giant launched an AI platform for human robots called General Robot 00 Technology (GR00T).
Nvidia sees robotics as unexplored space
According to a report from the Financial Times, the tech giant now positiones itself as the primary platform for robotic processing and AI-driven simulation training. Deepu Talla, vice president of robotics at NVIDIA, told the publication that the upcoming “Chatgpt moment” and robotics are coming.
However, this shift in focus could also mark the deprivation of AI chipsets and GPUs, the main driver of the company’s $3.6 trillion (about Rs 307.9 crore) valuation. The report claims that new areas of focus have been decided due to the recent increase in competition in the AI semiconductor space.
Chip makers such as AMD, Intel and Snapdragon have launched AI chipsets powered by Neural Processing Units (NPUs). Although the GPU space is still dominated by NVIDIA, entry into cloud computing giants such as Amazon, Google and Microsoft may further move away from their revenue.
According to the report, NVIDIA is now planning to position itself as a “full stack” solution for robotics, which includes hardware acceleration as well as training software. Recently, Jim Fan, head of NVIDIA’s embodied AI division, revealed that the company plans to train embodied AI agents (a AI agent equipped with sensors and mechanical parts) in simulations to prepare them for real-world challenges. .
NVIDIA is also reportedly preparing to launch the chipset Jetson Thor, built by the company’s ISAAC robotics platform, which provides its GR00T platform with a platform for simulation-based training. The latter also draws on the capabilities of generating AI to create a new way to train humanoid robots that do not require a lot of coding.
Talla told The Financial Times that the decision to focus on the robot market is based on the rise of generative AI models and the ability to train robots in simulated environments.
“In the past 12 months (this gap) has matured so well that we can now perform simulation experiments and use them in conjunction with generated AI, something we couldn’t do two years ago. We’re in order to make all of these companies Being able to accomplish these tasks provides a platform,” he added.
