Starcloud Launches Space-Based Data Centers with NVIDIA GPUs

Starcloud, a startup in NVIDIA's Inception program, is set to launch an AI-powered satellite equipped with NVIDIA H100 GPUs, marking a significant step toward creating sustainable data centers in space. The initiative aims to address the growing energy demands and cooling challenges of terrestrial data centers by leveraging the unique advantages of orbital infrastructure.

The Promise of Space-Based Data Centers

Starcloud envisions a 5-gigawatt orbital data center powered by expansive solar panels and cooling systems spanning approximately 4 kilometers. According to Philip Johnston, cofounder and CEO of Starcloud, space offers nearly unlimited, low-cost renewable energy, with environmental impacts limited to the launch phase. This approach could result in a 10x reduction in carbon dioxide emissions compared to Earth-based data centers.

NVIDIA H100 GPU's Cosmic Debut

The upcoming satellite launch in November will feature the NVIDIA H100 GPU, making it the first deployment of this advanced technology in space. The Starcloud-1 satellite, weighing 60 kilograms and roughly the size of a small refrigerator, is expected to deliver 100 times more GPU computing power than any previous space mission.

Sustainability Advantages

Space-based data centers offer unique sustainability benefits. Unlike Earth-bound facilities, Starcloud plans to use the vacuum of space as a natural heat sink, eliminating the need for freshwater cooling. Continuous solar exposure in orbit also removes the need for batteries or backup power, further enhancing efficiency. Johnston predicts that energy costs in space could be 10 times lower than on Earth, even when accounting for launch expenses.

Applications in Space

Early applications for space-based data centers include analyzing Earth observation data to improve crop monitoring and predicting local weather patterns. Real-time data processing in space can also enhance critical operations like wildfire detection and distress-signal response, reducing response times from hours to minutes.

Why NVIDIA?

Starcloud chose NVIDIA GPUs for their superior performance in training, fine-tuning, and inference, making them competitive with Earth-based data center workloads. The company plans to integrate the NVIDIA Blackwell platform in future launches, anticipating a tenfold increase in AI performance compared to the NVIDIA Hopper architecture. Additionally, Starcloud aims to run Gemma, an open model from Google, on H100 GPUs in orbit, demonstrating the viability of large language models in space.