Key Highlights

  • SpaceXAI will use Nvidia's new Vera chip to help its AI systems work faster, especially behind the scenes tasks like running code and organizing data.
  • SpaceXAI is expanding the computing power behind its Grok AI chatbot using Nvidia's technology, aiming for a massive scale of computing capacity.
  • SpaceXAI plans to send a version of this same Nvidia technology into space aboard its first AI satellite, called Starmind.
  • Nvidia's Vera chip is built with 88 processing cores and can complete tasks up to 1.8 times faster than typical computer chips.

A New Chip Built Specifically for AI That Takes Action

Nvidia (NASDAQ:NVDA) announced that SpaceXAI will use its new Vera chip to speed up the next generation of its AI systems. Nvidia describes Vera as the first chip of its kind built specifically to support AI "agents," AI programs that don't just answer questions but actually complete tasks like writing code, organizing data, and running simulations behind the scenes. SpaceXAI will use Vera to handle all of this background work faster, so its main AI chips can stay focused on the heavy lifting of actually running the AI models.

A senior Nvidia executive said that AI agents need computers that can take action quickly, not just generate answers, and that this new chip is built to give AI agents that speed. The head of SpaceXAI said the new chip gives the company enough computing power and speed to handle huge amounts of background work while keeping its main AI chips running at full capacity.

Built for Speed Behind the Scenes

Nvidia's Vera chip includes 88 processing cores and very fast memory, allowing it to complete tasks up to 1.8 times faster than standard computer chips used across the industry. It's designed to handle the kind of background work that AI agents constantly need done, like fetching information, running small programs, and organizing data, so the AI doesn't get stuck waiting.

From Earth to Orbit: A Plan to Put AI Computers in Space

SpaceXAI is using this same technology to expand computing power behind Grok, its AI chatbot, as it works toward an enormous scale of AI computing capacity. But the company isn't stopping on the ground. SpaceXAI's first AI satellite, called Starmind, is planned to carry a specially adapted version of this same Nvidia computing system into orbit.

Putting computers like this in space isn't simple. Power supply, heat management, and reliability all work very differently in orbit than they do in a Data Center on Earth. Nvidia and SpaceXAI are working together to adjust the technology for these space-specific challenges while keeping it built on the same core Nvidia system used on the ground.

Conclusion

This deal shows Nvidia's technology stretching further than ever, from speeding up an AI chatbot to potentially running computers in orbit. If SpaceXAI succeeds in adapting this system for space, it could be an early sign of a much larger shift toward putting AI computing power beyond Earth entirely. For now, the bigger and more immediate story is how much SpaceXAI leans on Nvidia to power the next version of its AI systems here on the ground.