Zheshang Securities released a research report stating that deploying data centers in low Earth orbit (LEO) has emerged as a promising new option in the future computing strategies of tech companies. China's LEO satellite constellations are approaching a peak launch period.
Private rocket companies are preparing to test new models such as Zhuque-3 and Tianlong-3, with launch costs expected to drop to ¥20,000/kg—close to SpaceX’s Falcon 9. Lower costs will enable more constellation operators to afford launches, unlocking latent demand. The rocket industry is currently in a supply-constrained capacity expansion phase, with high-value, high-barrier companies poised to benefit.
**Key Insights from Zheshang Securities:**
**Satellite Internet: Dual-Use Communication & Space Computing Potential** Satellite internet, formed by networked satellites, provides global broadband coverage for ground and aerial terminals. It strengthens military communication networks and expands terrestrial user services.
Space computing could address current data center limitations in power and cooling. Terrestrial data centers face power supply constraints, while space-based centers leverage uninterrupted solar energy—five times more efficient than on Earth—and eliminate water-cooling needs. Thus, LEO data centers present a strategic computing alternative.
**LEO Constellation Delays Drive Rocket Launch Demand Surge** China’s ambitious LEO constellation plans face urgency due to orbital slot and frequency allocation rules (ITU mandates 50% deployment within 12 years, full completion in 14 years). Current deployments, including GW and G60, lag at <1% of planned capacity.
Over the next five years, G60, GW, and other projects aim to launch ~16,000 satellites. Assuming 10 satellites per rocket, annual launches could surge from 54 in 2025 to 860 by 2030 (74% CAGR). By 2029, space computing may require 6,800 Chinese and 15,000 global launches.
IDC forecasts China’s computing power to grow at 49% CAGR (5457 EFLOPS by 2029), with global capacity reaching 14,130 EFLOPS (45% CAGR). If 2% of global computing shifts to space by 2029, China would need 6,800 launches, matching projected demand.
**Declining Launch Costs & Expanding Capacity** State-owned and private rocket firms are scaling up. For example: - **State Sector:** Aerospace Power’s new engine production base aims for 300 liquid engines/year by 2025. - **Private Sector:** Li Jian Super Factory targets 20 rockets/year; Tianbing’s Tianlong-3 aims for 30/year; iSpace’s Hyperbola-3 plans 20/year.
New models like Zhuque-3 and Tianlong-3 could cut costs to ¥20,000/kg, spurring demand.
**Key Industry Players:** - **Suppliers:** Hangyang (liquid oxygen fuel), Bright Laser (3D-printed parts), Sirui Materials (copper alloys), Sinomach Precision (bearings), SuperJolt (structural components), Gaohua Tech (sensors), Haoyue Holdings (critical systems). - **Potential OEMs:** Aerospace Power (CASC Sixth Academy), Aerospace Engineering (CASC First Academy), Zhongtian Rocket (CASC Fourth Academy).
**Risks:** Slower-than-expected progress in space computing, satellite internet deployment, or rocket capacity expansion.
Comments