Nvidia launches "space computing module", "space version" Vera Rubin to be released later.

```

NVIDIA is extending its AI computing footprint to Earth's orbit.

At the GTC annual developers conference held overnight, NVIDIA announced dedicated computing modules for space applications, unveiled a space version plan based on the Vera Rubin architecture, and released the enterprise-level AI agent platform NemoClaw, fully showcasing its ambition to expand in the field of AI infrastructure.

In his keynote speech, NVIDIA CEO Jensen Huang announced that six partners—Aetherflux, Axiom Space, Kepler Communications, Planet, Sophia Space, and Starcloud—would deploy NVIDIA computing hardware in orbit. According to the company, the new module targets scenarios such as orbital data centers, advanced geospatial intelligence processing, and autonomous space operations. Compared to the previously launched H100 GPU in space, the new module can deliver up to 25 times the AI inference performance.

The Vera Rubin space module will be officially launched at a later date. Currently, the IGX Thor and Jetson Orin products are available for shipment. At the same time, NVIDIA also released the enterprise-level AI agent platform NemoClaw, designed to provide enterprises with local AI agent deployment capabilities with security and privacy guarantees; it is currently in early Alpha testing.

Space Computing Layout: From H100 to Vera Rubin

NVIDIA’s foray into space computing can be traced back to November last year. According to Data Center Dynamics, Starcloud sent a NVIDIA H100 GPU into orbit on a test satellite in November last year, marking the first time a NVIDIA GPU had entered space.

The newly released space computing module system consists of three tiers: The Vera Rubin space module for high-intensity workloads such as orbital data centers, IGX Thor based on the Blackwell architecture for edge scenarios, and Jetson Orin for real-time processing of vision, navigation, and sensor data. Among these, IGX Thor and Jetson Orin are already officially available, while the specific launch date for the Vera Rubin space module has not yet been disclosed.

On the partner level, Aetherflux plans to use Vera Rubin for solar-powered high-performance AI inference in orbit; Sophia Space will adopt Jetson Orin in its modular platform for satellite operators; Kepler Communications has also chosen the Jetson series products.

In his speech, Jensen Huang noted: “As we deploy satellite constellations and carry out deeper space exploration, intelligence must exist wherever data is generated. AI processing capabilities must span both space and ground systems, transforming orbital data centers into exploration tools and spacecraft into autonomous navigation systems.”

It’s worth noting that earlier this month, NVIDIA posted a job opening for an orbital data center system architect. The listing describes it as “an opportunity to join an AI systems leader at the birth of an entirely new industry.”

Competitive Landscape: Google and Musk Join the Race

The space computing race is attracting multiple tech giants. According to Data Center Dynamics, Google plans to launch multiple batches of TPUs into space. Testing has already been completed via a particle accelerator that simulates the low Earth orbit radiation environment, with Planet involved in small-scale deployments. The long-term goal is to deliver gigawatt-level computing power to space.

SpaceX CEO Elon Musk also seeks to deploy an orbital AI data center “super constellation” with up to a million satellites. These satellites are reported to use Tesla’s proprietary chips, though the timeline remains unclear.

However, the commercial prospects of orbital data centers are not without skepticism. OpenAI CEO Sam Altman, short seller Jim Chanos, AWS CEO Matt Garman, and Gartner analysts have all expressed criticisms of the concept.

NemoClaw: Providing a Secure Foundation for Enterprise AI Agents

In the field of enterprise AI agents, NVIDIA simultaneously launched the NemoClaw platform. According to TechCrunch, NemoClaw is built on the open-source local AI agent project OpenClaw, with NVIDIA collaborating with OpenClaw founder Peter Steinberger to develop this enterprise version.

NemoClaw’s core positioning is to layer enterprise-level security and privacy mechanisms atop OpenClaw, allowing companies to access the platform with a single command and enforce unified control over agent behavior and data handling. The platform supports access to any programmatic agent or open-source AI model, including NVIDIA’s own NemoTron open-source models, and is deeply integrated with NVIDIA's NeMo AI agent software suite. Importantly, the platform remains open and hardware-agnostic, not requiring NVIDIA GPUs to run.

Jensen Huang compared the advent of OpenClaw to historic moments like Linux, Kubernetes, and HTML, stating: “Every company today needs to develop an OpenClaw strategy, a system strategy for agents.”

Currently, NemoClaw is still in the early Alpha stage. In developer documentation on its website, NVIDIA acknowledges that “rough edges are expected” and that it is working toward production-ready sandbox orchestration capabilities. Enterprise AI agent platform competition is heating up—OpenAI launched its OpenAI Frontier enterprise platform in February this year, and Gartner identified AI agent governance platforms as a key infrastructure for enterprise adoption of AI technologies in a report last December.

Risk warning and disclaimerMarkets involve risk, and investment should be approached with caution. This article does not constitute personal investment advice, nor does it take into account an individual user’s specific investment goals, financial situation, or needs. Users should consider whether any opinions, viewpoints, or conclusions herein are suitable for their circumstances. You bear all responsibility for any investment actions based on this article. ```