Jensen Huang: Orbital Data Centers' Economics Remain Poor Amid Cooling Hurdles

Nvidia CEO Jensen Huang addressed questions about the viability of orbital data centers during the company's fourth-quarter fiscal 2026 earnings conference call held February 25, 2026.

Credit: Lee Jin-man - AP

Huang responded to an inquiry from analyst Antoine Chkaiban, who asked about the feasibility, timeline, and economic aspects of space-based data centers that some Nvidia customers have explored.

Jensen Huang outlined the current limitations and potential future developments for such infrastructure, emphasizing differences in energy availability, spatial constraints, and thermal management between terrestrial and orbital environments.

Huang described space as offering significant advantages in terms of energy and expansion potential.

"As you know, the way that space works is radically different than how it works down here. There's an abundance of energy, but solar panels are large, but there's plenty of space in space," Jensen Huang said.

He pointed out that solar power accessibility in orbit eliminates many of the energy sourcing issues faced on Earth, where data centers compete for grid capacity and renewable integration.

Yet Huang stressed that these benefits do not yet offset the overall costs, which include launch expenses, specialized hardware design, and ongoing operational demands unique to vacuum conditions.

Cooling emerged as the central technical obstacle in Jensen Huang's remarks.

"The heat dissipation, it's cold in space, however, there's no airflow, and so the only way to dissipate heat is through conduction, and the radiators that you need to create are fairly large," he explained.

He detailed how the absence of atmosphere prevents convective cooling, forcing reliance on radiative systems that require extensive surface area for heat rejection.

This setup increases mass and complexity, driving up deployment costs via higher launch payloads.

Jensen Huang contrasted this with Earth-based methods, noting that standard liquid cooling systems prove impractical in orbit.

"Liquid cooling is obviously out of the question because it's heavy and, you know, freezes. The methods that we use here on Earth are a little different than the way we would do it in space," he added.

Such adaptations demand engineering innovations in materials and architecture to handle extreme temperature swings and microgravity effects on hardware reliability.

Despite these challenges, Huang affirmed Nvidia's active involvement in space computing. He revealed that the company has deployed graphics processing units beyond Earth's atmosphere.

"NVIDIA is already the world's first GPU in space, Hopper is in space," Jensen Huang stated.

The Hopper architecture, part of Nvidia's data center lineup, supports tasks in orbital settings where low-latency processing is critical. Huang highlighted specific applications that benefit from on-site computation, particularly in remote sensing and data handling.

Huang elaborated on practical use cases for GPUs in space, focusing on high-resolution imaging enhanced by artificial intelligence.

"One of the best use cases of GPUs in space is imaging. To be able to image at extremely high resolutions using of course optics and artificial intelligence, and to be able to do that computation of reprojection of different angles and be able to up res and do noise reduction, and just be able to see, be able to image at very large, very high resolutions, extremely large scales, and very, very fast," he said.

This capability allows satellites to process vast datasets in real time, applying AI for tasks like angle reprojection, resolution enhancement, and noise filtering without transmitting raw data back to ground stations.

Huang further explained the efficiency gains for space-based AI operations.

"It's hard to do that, you know, by sending petabytes and petabytes of imaging data back here on Earth and doing that work. It's easier just to do it out in space. Ignore all of the data collected and processed until you see something interesting," he noted.

By filtering irrelevant information at the source, orbital systems reduce bandwidth needs and enable faster decision-making for applications in Earth observation, disaster monitoring, and scientific research.

Jensen Huang indicated that these scenarios demonstrate viable niches for space computing even as broader data center deployments face hurdles.

Huang projected gradual advancements in the field.

"Well, the economics are poor today, but it's gonna improve over time," he said.

Factors such as reductions in launch costs through reusable rocketry, improvements in radiator efficiency, and scaled production of space-hardened components could shift the cost-benefit analysis.

Jensen Huang's comments came amid Nvidia's report of record quarterly revenue of $68.1 billion for the period ended January 25, 2026, driven largely by data center demand for AI accelerators.