NVIDIA’s Space 1 Vera Rubin Brings Data Center Computing to Orbit

March 25, 2026

Woodrow Bellamy III

NVIDIA unveiled the new Space-1 Vera Rubin Module during its GTC event. (Image: NVIDIA)

NVIDIA CEO Jensen Huang announced the launch of the Space‑1 Vera Rubin Module during his keynote at the company’s 2026 GPU Technology Conference (GTC) last week.

The Space‑1 Vera Rubin Module is the space-qualified variant of NVIDIA’s new Rubin computing platform and comprises six new chips named in honor of American astronomer Vera Florence Cooper Rubin. First announced at CES 2026, the platform includes the NVIDIA Vera CPU, NVIDIA Rubin GPU, NVIDIA NVLink 6 Switch, NVIDIA ConnectX‑9 SuperNIC, NVIDIA BlueField‑4 DPU, and NVIDIA Spectrum‑6 Ethernet Switch. Together, these components are designed to deliver compute, networking, and storage/infrastructure functions directly within artificial intelligence (AI) data centers, with the Space-1 Vera Rubin Module now extending that capability to orbit.

NVIDIA provided the following technical breakdown of the Rubin platform components:

NVIDIA Vera CPU: 88 custom-designed NVIDIA Olympus cores optimized for next-generation AI factories, with full Arm compatibility.
NVIDIA Rubin GPU: High-performance AI compute featuring HBM4 memory and the new NVIDIA Transformer Engine.
NVIDIA NVLink™ 6 Switch: Sixth-generation scale-up interconnect fabric delivering 3.6 TB/s of GPU-to-GPU bandwidth.
NVIDIA ConnectX®‑9 SuperNIC: High-throughput, low-latency endpoint networking for large-scale, distributed AI workloads.
NVIDIA BlueField®‑4 Data Processing Unit (DPU): A dual-die architecture combining:

A 64-core NVIDIA Grace CPU for infrastructure offload and security functions.
An integrated NVIDIA ConnectX‑9 high-speed networking chip for tightly coupled data movement.

NVIDIA Spectrum™‑6 Ethernet Switch: Scale-out networking using co-packaged optics to improve efficiency and reliability.

Upon announcing the Space‑1 Vera Rubin Module, NVIDIA emphasized the platform’s ability to push advanced AI computing directly to space-based edge platforms, enabling applications such as orbital data centers, geospatial intelligence, and what the company describes as “autonomous space operations.”

NVIDIA released this image with its announcement of the new Space-1 Vera Rubin Module. (Image: NVIDIA)

“We’re going to space, and we’ve already been in space. Thor is radiation approved, and we’re in satellites,” Huang said during his keynote speech. “In the future, we’ll also build data centers in space — obviously very complicated to do so.”

Several commercial space companies are already deploying NVIDIA computing platforms in active space missions, including Aetherflux , Axiom Space, Kepler Communications, Planet, Sophia Space, and Starcloud.

Kepler Communications, a commercial satellite communications provider, shared an update on its use of NVIDIA computing across its Tranche 1 optical data relay constellation, which the company describes as the “world’s first commercially operational optical data relay network.” The Kepler Network is designed to support optical connectivity, distributed on-orbit computing, and secure payload hosting.

Applications enabled by the Kepler Network include “AI-driven Earth observation analytics, multi-sensor data fusion, RF signal intelligence, autonomous network operations, and intelligent data optimization executed directly within the network.” According to Kepler, its initial on-orbit compute capability is supported by 40 NVIDIA Jetson Orin modules, “deployed as distributed edge compute GPUs across 10 satellites and interconnected through Kepler’s real-time optical communications network.”

“This architecture removes one of the longest-standing constraints in space operations,” said Mina Mitry, CEO of Kepler Communications. “By leveraging NVIDIA AI infrastructure in our optical network, data can be processed, routed, and acted on in orbit rather than waiting to return to Earth. As we extend the scale of our infrastructure, this becomes a natural extension of terrestrial computing—enabling AI-driven detection, faster decision-making, increased resilience, and entirely new mission architectures for our customers and partners.”

Planet Labs, the satellite imagery and geospatial intelligence provider founded by three NASA scientists, is also leveraging NVIDIA platforms, including Blackwell and IGX Thor, to process imagery directly on orbit. This approach reduces processing time from several hours to just seconds, converting raw pixel data into analysis-ready insights. Planet is also working with NVIDIA to deploy Blackwell GPUs within its satellite data ground-processing systems.

“Planet images the Earth every day, a data challenge that requires the world’s most advanced computing,” said Will Marshall, CEO of Planet. “By integrating NVIDIA’s accelerated platform from space to ground, we are supercharging our ability to index the physical world. Using NVIDIA CorrDiff AI models, we are moving from raw pixels to actionable insights in near real time. Together, we are enabling a revolutionary leap in planetary intelligence, helping humanity make smarter decisions at the speed of global change.”

Across GTC, other commercial space companies echoed a similar theme: pushing the latest commercial off the shelf (COTS) embedded processing to the edge of satellites and spacecraft. Local processing of complex sensor data, according to NVIDIA, reduces downlink requirements, decreases latency, and lowers the cost and time associated with transmitting large volumes of raw data to Earth for ground-based analysis.

Huang also addressed the unique challenges of deploying high-performance computing platforms like the Space‑1 Vera Rubin Module in the space environment.

“In space, there’s no conduction, there’s no convection—there’s just radiation,” Huang said. “So we have to figure out how to cool these systems in space. But we’ve got a lot of great engineers working on it.”

NVIDIA expects the Space‑1 Vera Rubin Module to become available later this year.