China Tests AI Computing in Orbit, Challenging Earth's Data Center Dominance
Baca dalam 60 detik
- China meluncurkan satelit Supercomputing-1 yang mampu memproses data pengamatan Bumi langsung di antariksa, memangkas waktu pemrosesan dari jam ke menit.
- Langkah ini merupakan bagian dari strategi besar China membangun jaringan komputasi terintegrasi yang melibatkan industri roket, satelit, semikonduktor, dan AI.
- Inisiatif ini berpotensi mengubah lanskap infrastruktur digital global dan membuka peluang kolaborasi teknologi bagi negara-negara berkembang, termasuk Indonesia.
China has marked a new milestone in the space computing race by launching the Supercomputing-1 satellite, designed to process Earth observation data directly in orbit rather than sending it to ground-based data centers. The satellite is part of nine payloads launched using the Kinetica 1 Y18 rocket, operated by commercial launch provider CAS Space, on 20 September.
The satellite, developed by Chinese AI computing infrastructure company S-AIDC, carries a high-resolution optical payload and an AI image-processing computer. With that capability, the satellite can independently identify objects such as forest fires, storms, floods, or ships, and then send only the relevant analysis results to ground stations. This approach is considered more efficient than transmitting large volumes of raw data.
China's move is not merely a technology experiment. In early June, the Chinese government approved the establishment of the Space Computing Industry Innovation Center, an initiative that brings together rocket and satellite manufacturers, semiconductor plants, and AI technology companies to build a space computing network. The goal is clear: to reduce dependence on ground-based data centers that are increasingly strained by the explosion in AI demand.
Pressure on ground infrastructure is the main trigger. AI data centers housing hundreds of thousands of accelerators consume unprecedented amounts of electricity, straining power grids and triggering public opposition over land, cooling water, and noise issues. According to reports, in the second quarter of 2026, data center projects worth US$68 billion were delayed due to community resistance.
Space offers an attractive alternative: unlimited land, access to solar energy, constant cold temperatures, and silent vacuum. SpaceX, through its AI1 satellite, plans to provide 150 kW of peak computing power and build an 11 million-square-foot Gigasat facility to produce 6,000 satellites in order to reach a target of 1 GW per year by 2027. Nevertheless, China's approach with a single edge AI satellite is considered more practical and applicable for current needs.
"Launching a single satellite with edge AI capability is indeed far from a full orbital data center, but China's approach could be a more practical and immediate application for advanced computing in space," according to an analysis by Digitimes.
For Indonesia, this development opens both opportunities and challenges. As an archipelagic country prone to disasters, the ability to process Earth observation data in real time in orbit could accelerate emergency response, weather monitoring, and natural resource management. However, Indonesia still depends on foreign technology and does not yet have its own space computing infrastructure. Cooperation with China or SpaceX could be a shortcut, but it also creates strategic dependence.
The question is, will Indonesia merely become a market for orbital computing technology, or will it begin building its own capacity? With a strategic geographic position and continuously growing data needs, the adoption of edge computing on satellites could be a major leap—provided there is political will and serious research investment.
