Optimistic about space data centers, Goldman Sachs sharply raises its low-Earth-orbit satellite deployment forecast: by 2031, it is set to jump to 300k satellites, with the optimistic scenario reaching 400k.

From rocket manufacturing, satellite hardware, to ground terminal equipment—everything is standing at a turning point of historical-scale expansion.

According to the Qiu Feng Trading Station, Goldman Sachs released its latest global technology research report on Tuesday, substantially raising its 2031 low Earth orbit (LEO) satellite benchmark installation capacity forecast from 42k satellites to 305k satellites, an increase of more than 634%. This rare-scale revision signals a re-pricing of the LEO satellite industry’s scale by Wall Street.

The core logic behind the upward revision is not a linear extension of satellite internet demand, but the inclusion of an entirely new application scenario: space data centers.

Goldman Sachs expects that after 2029, space data centers will replace satellite internet and become the main engine driving global LEO installation capacity growth; their share will rise from zero to 63%, and further to 79% in 2031. Meanwhile, the large-scale entry of Chinese satellite operators is also listed as a structural variable that cannot be ignored.

Major forecast upward revision: after 2029, installation capacity surges exponentially

Goldman Sachs first released its LEO satellite industry forecast in February 2025.

The report shows that by end-2025, the actual global LEO satellite installation capacity is about 9,982 satellites, slightly above the previously forecast 9,700, with launch progress already ahead of initial expectations.

The divergence between the new and old forecasts is not significant in the near term, but it expands sharply over time.

The new forecast for 2026 is 13,088 satellites, only about 9% higher than the old forecast of 12,000; the new forecast for 2028 is 23,796 satellites, about 25% higher than the old forecast of 19,000. The real gap appears in 2029—new forecasts jump to 90,042 satellites, with a year-over-year increase of 278%; in 2030 it further rises to 164,244 satellites; and in 2031 it reaches 305,293 satellites, which is exactly 634% higher than the old forecast of 42,000.

Goldman Sachs clearly points out that the huge difference between the new and old forecasts stems from incorporating space data centers into the satellite application scenario forecast framework. This newly added dimension fundamentally changes the industry’s installation capacity growth curve shape.

Space data centers: the largest incremental growth engine after 2029

Broken down by application scenario, Goldman Sachs’ forecast reveals a key structural shift.

From 2025 to 2028, the incremental increase in global LEO installation capacity is 100% from satellite internet; entering 2029, the share of space data centers jumps to 63%, while satellite internet falls to 37%; by 2031, space data centers further rise to 79%. Under the benchmark scenario, installation capacity is 241,486 satellites, while satellite internet installation capacity is only 63,807 satellites.

Goldman Sachs explains in the report the core appeal of space data centers: the ability to access low-cost solar power without limit, plus edge-computing capability that can directly process satellite data.

However, Goldman Sachs also admits that the technical feasibility of space data centers still needs to be validated. This means whether the installation capacity forecast after 2029 can be realized depends, to a considerable extent, on whether this new business model can make the leap from concept to large-scale implementation—making it the biggest uncertainty factor in the entire forecast.

Blue Sky scenario: if China’s constellations proceed as planned, 2031 could reach 395,624 satellites

Goldman Sachs also proposed an optimistic “Blue Sky scenario,” with key assumptions including: rocket capacity performing better than expected, acceleration in the commercialization of satellite communications, and the long-term realization of China’s submitted 200k-satellite LEO constellation plan.

Under the Blue Sky scenario, global LEO installation capacity in 2031 can reach 395,624 satellites, about 30% higher than the benchmark scenario’s 305,293 satellites.

The biggest difference between the two scenarios comes from China: in the benchmark scenario, China’s 2031 installation capacity is 23,750 satellites, accounting for 8% of the global total; in the Blue Sky scenario, this figure jumps to 101,148 satellites, accounting for 26% globally.

As of end-2025, China’s satellite operators currently have only 253 satellites actually operating in orbit, but they have collectively submitted applications for more than 200k medium- and low-orbit satellites to the International Telecommunication Union (ITU).

Goldman Sachs interprets this as a strategic move to secure spectrum and orbital resources, rather than a launch plan that must be realized in the short term.

Rocket technology: the core variable determining whether installation capacity forecasts can be delivered

Goldman Sachs lists progress in rocket technology as the key constraint on whether the overall forecast can materialize, and analyzes it across three dimensions: launch frequency, payload capacity improvement, and reusable technology.

In terms of launch frequency, leading operator Starlink already achieved an average launch cadence of once every three days in 2025; in April 2026, SpaceX set another record by continuously launching two rockets within 19 hours, showing there is still room for further efficiency improvements.

In terms of payload capacity, the current main reusable rockets’ LEO payload is about 17,500 kilograms; the next generation rockets’ payload will reach 100 to 150 tons, implying a payload capacity improvement of about 6 to 9 times.

For China, breakthroughs in domestic reusable rockets are a key milestone for accelerating the expansion of LEO satellites. Current progress includes:

• Long March 10 (LM-10): In February 2026, it successfully completed a first-stage soft landing and sea splashdown test; on July 10 it completed first-stage rocket net recovery;

• Zhuque-3: In December 2025, testing of parts of a flight was successful; in the second quarter of 2026 it will conduct a recovery test flight;

• Hyperbola-3 (Double Curves 3): planned to complete its first launch by the end of 2026;

• PALLAS-1: a reusable rocket by Galaxi Aerospace; planned to complete its first launch in 2026.

The above highlights come from the Qiu Feng Trading Station.

For more detailed interpretations, including real-time commentary, frontline research, and more, please join 【**Qiu Feng Trading Station ▪ Annual Membership**】

![](https://img-cdn.gateio.im/social/moments-8293f7dff1-0340ba37ce-8b7abd-7649e1)

Risk warning and disclaimer terms

          

            The market is risky; investment requires caution. This article does not constitute personal investment advice, and it does not consider any individual user’s special investment objectives, financial situation, or needs. Users should consider whether any opinions, viewpoints, or conclusions in this article align with their specific circumstances. Investing based on this is at your own risk and responsibility.