Bench Talk

(Source: TON_PHOTO/stock.adobe.com)

“Every software engineer, every creative artist, everybody who uses computers today as a tool will need an AI supercomputer.” — Jensen Huang, CEO, NVIDIA

It's no longer just researchers or enterprises that require access to artificial intelligence (AI).

That was the sentiment NVIDIA CEO Jensen Huang wanted to convey in his keynote address at CES 2025.

But how can engineers, small startups, or independent creators access the power of AI when, traditionally, so many barriers to adoption exist? Not surprisingly, a recent US Census Bureau report revealed that AI use increases with firm size, meaning larger firms are more likely to adopt AI technology.^[1]

High costs, complex infrastructure, and learning curves often stand in the way of the everyday professional who wants to tap into the power of AI. When Huang revealed NVIDIA's latest innovations at CES 2025, he didn't just show off some new tools and hardware. He shared a tangible way to make AI accessible to even the individual creator with tools like Project DIGITS, NEMO, and Blackwell GPUs. Huang's vision is more than technology; it's about empowering anyone with the desire to innovate and equipping everyone with the full potential of AI.

AI Tools for Everyone

In order to democratize AI, technologies need to be easy to access, affordable to use, simplistic enough with user-friendly interfaces, and have decent resources for people to learn. By limiting AI technology to large research firms and enterprises, we limit the level of innovation and economic growth in our society. At CES, Huang successfully relayed NVIDIA’s plans for democratizing AI, explaining that making AI accessible to all means simplifying workflows and lowering the barrier to entry. Until now, AI development required a high level of expertise and powerful hardware. However, with ready-to-use solutions like NEMO, Omniverse, and Cosmos, AI is more accessible than ever.

NEMO gives developers pre-trained models and intuitive fine-tuning capabilities to easily adapt AI for specific tasks. For example, a healthcare startup can employ NEMO to build an AI model that analyzes patient records without starting from scratch. This type of plug-and-play option reduces development time and eliminates the technical complexity.

Omniverse allows engineers to collaborate on digital twins, model real-time environments, and integrate generative AI into design workflows. With Omniverse, an automotive manufacturer can test vehicle designs in a simulated environment to save time and resources that would be wasted building physical prototypes.

Cosmos leans into robotics and autonomous systems, enabling physical AI that teaches machines to understand and interact with the real world. Robotics companies, for example, can train AI to recognize objects and adjust routes in real time. Collectively, these tools allow more professionals to harness AI in their efforts to innovate and solve challenges across industries.

Use cases abound for AI’s emerging capabilities. Engineers can now design digital twins of manufacturing systems to optimize performance before physical implementation. Similarly, developers can reduce project timelines by leveraging ready-made AI blueprints. These tools can also work together so an engineer can use Omniverse to create a digital twin, train its AI models using NEMO, and then test its physical interactions through Cosmos. CES put on full display just how NVIDIA’s suite of tools addresses all stages of AI development and empowers anyone to create and innovate, regardless of their technical background, in order to make AI workflows more accessible.

Project DIGITS Puts AI in Your Hands

CES also featured Huang discussing Project DIGITS, a wireless, compact AI supercomputer designed to fit on a desk while delivering the power of a traditional supercomputing setup. DIGITS was built on the GB110 chip, combining NVIDIA’s Blackwell GPU and Grace CPU technologies. Its portability and scalability make it a great solution for engineers, creators, and developers. DIGITS runs NVIDIA’s full AI stack, which includes tools like Omniverse and NEMO. It’s compatible with PCs, Macs, and Linux systems for maximum accessibility, and it reduces costs and setup time, providing AI capabilities without the need for traditional supercomputing infrastructure. Putting AI supercomputing in the hands of individuals and small teams allows for seemingly limitless advancements in engineering, design, and beyond.

Blackwell GPUs: Powering AI for All

What lies at the core of these NVIDIA AI advancements is the Blackwell GPU, which powers tools like DIGITS. Blackwell GPUs deliver outstanding performance while minimizing computational overhead, which makes AI applications efficient and accessible. With up to 4 petaflops of AI performance and 380 teraflops of ray tracing power, these GPUs enable real-time neural rendering and advanced CAD simulations. Their deep learning super sampling (DLSS) technology enhances rendering speeds while preserving high resolution and detail. For engineers, Blackwell GPUs accelerate rapid prototyping and real-time design visualization, as well as streamline workflows in manufacturing and robotics. Creators can use these GPUs to optimize generative tasks like video production, graphic design, and animation. Blackwell GPUs are what DIGITS is built on to create an environment where hardware and software can come together to democratize AI for a variety of users.

Conclusion

After hearing Huang’s vision and watching the advancements presented at CES, we can see that NVIDIA has clear ambitions to make AI more accessible than ever, giving individuals and teams the power to create, innovate, and push boundaries. But beyond a vision, they're making it happen with tools like Project DIGITS, powered by Blackwell GPUs so that engineers, creators, and developers can overcome traditional barriers and employ AI’s full potential.

When the keynote address at CES concluded, audiences everywhere left with a glimpse into a revolutionary future. Whether designing smarter systems, streamlining workflows, or exploring new ideas, NVIDIA’s latest technologies bring the power of AI to everyone.

Sources

[1]  https://www.census.gov/newsroom/blogs/research-matters/2024/12/ai-use-small-businesses.html