NVIDIA might be the best-positioned company to help grow the metaverse alongside giants like Meta, Google, and Amazon. NVIDIA’s core competency, producing powerful graphics hardware with chips capable of running huge calculations in parallel, is the same stuff that allows massive metaverse projects to run.
More technological advancements and more massive scale is coming to the metaverse very soon and much of the necessary innovation is likely to touch NVIDIA, but not just its hardware. NVIDIA has its own metaverse project called Omniverse that acts as a sort of toolkit to bring together all the components that make things work.
What Is the Metaverse?
To give more context, the metaverse is a bit of a catchall term for collaborative, online, open-design environments. These types of environments aren’t exactly new; people have been meeting, interacting, and creating content in these spaces for decades.
Blocky graphics and stilted motions defined early projects like Second Life, a metaverse game that came out in 2003, but those pixelated graphics have now graduated to high-definition, Hollywood-style, 3D computer animations with AI-guided, humanlike movements.
To join a metaverse you need an avatar, or digital representation of yourself. It’s a lot like a video game where users choose from among a static selection of characters to play as. In the metaverse, things can be taken a step further, users are able to create their avatar without restriction, often on design software like Adobe Illustrator.
Metaverse designers can also create different artifacts, like a chair or a flower, or a virtual house with custom-made backgrounds, or an amusement park. A person in the metaverse uses the avatar they created to go and meet people online and interact with them in a way that they’ve defined. They can use artifacts for trade.
Today’s Metaverse
Creating for today’s metaverse requires talent and big computing power. Unlike the blocky floating characters of 20 years ago, designers have the capability to make avatars that move like humans that are modeled in 3D with the quality of a Pixar movie (the Pixar technology is what NVIDIA uses in its Omniverse project). So even if an avatar is a blocky LEGO character, that LEGO character moves around like a human. If it’s a horse, it moves like a horse. It can have facial expressions. It can speak in any language. The designer determines how their avatar is perceived by others.
Applications of the Metaverse
The applications are nearly limitless and go beyond social and entertainment. The same tools can be used in engineering and fabrication scenarios. A car company, for instance, can move away from designing automobile modules in a piecemeal fashion and have all their teams collaborate in the metaverse. Companies can dynamically and collaboratively design a car, or a factory, or train robots to be more efficient in the real world. People from around the world can be in the same office, the same boardroom in the metaverse.
The computing advancements that underpin this better-looking, better-performing world — like ray tracing, posture estimation, Universal Scene Description — are complex and expensive. But all of them come down to two issues: rendering 3D graphics and training and implementing AI.
For a long time, NVIDIA has been an innovator in graphical processing units (GPUs). Their chips are designed to perform massive calculations for a computer’s central processing unit (CPU). NVIDIA chips contain several thousand cores and can perform 5,000 or 6,000 calculations in parallel. That’s about 100 times more calculations at the same time than some chip-making competitors like Intel, and it means exponentially better graphics and more fuel for calculation-heavy AI.
When NVIDIA made Omniverse and combined the hardware with a full stack of developer software, it unlocked potential for the metaverse to grow in new ways. The NVIDIA Omniverse offers AI in the stack and the hardware capability to run AI. More important, because Omniverse is software agnostic in a lot of ways, huge players like Amazon, Meta, and Google, which have much bigger AI units but maybe not the same level of graphics, will be able to use it to push huge structural improvements in their own metaverse projects.
But the tool is not reserved for big companies. Omniverse is open, very flexible, and accommodates multiple design capabilities. Collaboration means that different people must be able to incorporate their own creations into that environment and use it how they want.
For example, because the Omniverse can be viewed as one of the biggest supercomputers available, it’s attractive to the scientific community. Researchers could use NVIDIA’s Omniverse to run climate modeling scenarios or quantum physics, which require large-scale calculations and open collaboration.
But perhaps the most important constituency is developers, who will use Omniverse more like a software stack to literally design the look and feel of the metaverse at large. They’ll get access to libraries or design environments along with a preloaded AI or graphic stack that they can use natively.
NVIDIA expects that a majority of designers will use whichever tools they already have and just bring artwork and artifacts into Omniverse using connectors or import-export. NVIDIA will give designers some tools while fostering an environment where they can design things without being restricted by whatever NVIDIA is offering. NVIDIA doesn’t expect people to just sit on the Omniverse.
NVIDIA and the Metaverse
NVIDIA will probably maintain a free version of Omniverse that makes sense mostly for individual designers and developers. A lot of the NVIDIA software-hardware stack is not always cheap, so a lot of individual designers might hesitate to use it. A free version allows those doing gigs, or working from home, to operate on NVIDIA’s software and hardware stack.
All these elements — the connectivity of Omniverse tools, the ability to work with large companies and individual designers, the power of the NVIDIA hardware — means that NVIDIA is poised to have keystone position as the metaverse grows.
About Alex Ermolaev
Alex Ermolaev is currently an Advisor at VasoGnosis. Before this, he was Managing Director at TeamCamp. Prior to this role, he was employed at Change Healthcare Inc., holding the title of Director of AI. He has also held roles at Nvidia Corporation, Alanta, Ergo Growth Partners, and Microsoft. He has 20 years of experience building bleeding-edge software businesses.
This technology industry article was adapted from the GLG Roundtable “Nvidia’s Omniverse & the Metaverse: The Importance of AI and the Focus on Developers.” If you would like access to events like this or would like to speak with technology industry experts Alex Ermolaev or any of our approximately 1 million industry experts, please contact us.