At the 2025 Consumer Electronics Show (CES) held in Las Vegas, Nvidia unveiled a series of innovative products and partnerships that reflect the company's commitment to advancing artificial intelligence (AI) across various sectors. Highlighting its ongoing efforts to revolutionise the AI landscape, Nvidia's announcements encompassed new hardware, such as a personal AI supercomputer named Digits, as well as groundbreaking software solutions targeting the automotive and robotics industries.
Nvidia's Digits supercomputer, which will be available for purchase starting in May for approximately $3,000, aims to empower data scientists and AI enthusiasts by providing them a powerful computing resource right on their desks. According to Jensen Huang, founder and CEO of Nvidia, "Placing an AI supercomputer on the desks of every data scientist, AI researcher and student empowers them to engage and shape the age of AI." The Digits machine, compact enough to fit on a desk, integrates the GB10 Grace Blackwell superchip tailored for optimal AI computations. With up to 128GB of unified memory and 4TB of NVMe storage, Digits can execute large language models with up to 200 billion parameters, allowing users to experiment with models close to the capabilities of commercial giants like OpenAI’s GPT-4.
During the same event, Nvidia introduced its "Cosmos" platform, which enables researchers and developers to create extensive synthetic datasets required for training AI systems, particularly in developing advanced robotics and autonomous vehicles (AVs). Rev Lebaredian, Nvidia's vice president of Omniverse and simulation technology, explained, “Today's humanoid developers have hundreds of human operators performing thousands of repetitive demonstrations just to teach a couple of skills.” He elaborated that Cosmos allows for rapid generation of artificial video footage grounded in real-world physics, thus improving the efficiency of AI training processes. The platform employs a method known as multiverse simulation, synthesising numerous possible outcomes to aid in the training of robotic systems and AVs.
On the automotive front, Nvidia announced a significant partnership with Toyota, which will utilise Nvidia’s Drive AGX Orin supercomputer and the DriveOS operating system to enhance its next-generation vehicles with automated driving capabilities. Ali Kani, Nvidia's vice president of automotive, noted, “Toyota is actually a great example of our cloud-to-car strategy.” This partnership marks a continuation of Nvidia's collaboration with Toyota, which has been ongoing since 2019, encompassing a range of AI-driven projects.
The autonomous vehicle technology sector is set to benefit not only from this collaboration but also from Nvidia's expansive suite of AI platforms. Huang asserted that the self-driving cars sector could emerge as the "first trillion-dollar robotics market," highlighting its vast commercial potential. Alongside Toyota, other industry players such as Aurora Innovation and Continental are also integrating Nvidia’s technology, aiming to deploy driverless trucks powered by the Nvidia Drive Thor system-on-a-chip.
In conjunction with these announcements, Huang drew parallels to the rise of generative AI models, suggesting that the new technologies could help propel advancements in robotics similarly to how large language models have transformed enterprise AI. “The ChatGPT moment for robotics is coming,” he remarked, indicating a growing synergy between AI models and robotic applications through platforms like Cosmos, designed for extensive training scenarios using synthetic datasets.
These developments underscore Nvidia's strategic positioning as a key facilitator of AI's integration into diverse industries, reflecting a robust pipeline of innovations designed to enhance functionality, efficiency, and engagement across research and commercial domains.
Source: Noah Wire Services