CFO says SpaceX (SPCX.US) has increased confidence in its 100 billion dollar annualized revenue target! Terrestrial computing power takes big orders, and the space AI vision opens up new horizons

The Zhitong Finance App learned that new developments revealed by SpaceX (SPCX.US), an AI and space exploration leader founded and led by Musk at the much-anticipated Goldman Sachs Communacopia+ Technology Conference, that the newly signed AI computing power infrastructure resource supply agreement will bring in monthly revenue of US$1.11 billion from December, about US$13.32 billion in full 12 months, enhancing management's confidence to achieve the annual revenue operating rate target of 100 billion US dollars. It also highlights that the AI computing power infrastructure-related business is rapidly becoming the company The core growth engine. SpaceX Chief Financial Officer Brett Johnson also announced the company's important dynamic schedule for several growing business lines, such as the expansion of terrestrial computing power, deployment of AI computing power in space orbit, large-scale commercialization of Starships, and direct satellite mobile phones.

$100 billion is SpaceX's goal of reaching an annualized revenue run rate (annual revenue run rate) scale by the end of 2026. The key to this incredibly strong growth target is the rate of revenue growth: to reach this target, SpaceX requires monthly revenue of about US$8.333 billion, which is about 3.2 times the company's monthly average of US$2.6 billion in the second quarter.

Musk previously posted on social media X predicting that SpaceX's annual revenue will reach about 3.5 trillion US dollars around 2033, which is more aggressive than the “1 trillion US dollars by 2030” proposed earlier; if Wall Street's estimated revenue of 100 billion US dollars in 2027 is used as a base, it is equivalent to a 35-fold increase over six years, with a compound annual growth rate of about 81%. However, this aggressive prediction is Musk's personal prediction and is not an official company guideline.

The computing power agreements between Google and Anthropic correspond to approximately US$920 million and US$1.25 billion per month, respectively. If the new contract starts as scheduled and the first two businesses maintain this scale, the total of the three projects will be approximately US$3.28 billion per month. These figures show that external customer purchases are becoming an important support for SpaceX's revenue expansion. However, the $6.7 billion AI cloud computing service agreement revenue prospects lack a clear contract period in the current SpaceX fundamental expansion materials, and cannot be directly added to the calculation as monthly or annual revenue.

These positive developments can be described as supporting the further extension of SpaceX's growth sources to commercial computing power services, while reserving long-term upward space for orbital computing—that is, ground computing power contracts support recent revenue expansion, while the space AI data center vision is actively expanding the room for long-term growth. For AI computing power-themed transactions, large AI cloud service procurement agreements led by SpaceX help to significantly strengthen the market's optimistic demand outlook for the AI computing power industry chain.

At the Goldman Sachs CommunaCopia+ technology conference, SpaceX showed strong confidence that it will achieve an annualized operating rate of 100 billion US dollars. The company also plans to expand the terrestrial AI infrastructure capacity from slightly more than 2 gigawatts by the end of 2026 to 5 to 10 gigawatts in 2027, and explore orbital computing to break through ground-based power constraints. SpaceX plans to begin commercializing modular orbital computing power at the end of this decade. The long-term goal is to deploy 100 gigawatts of AI computing capacity to orbit every year — if operated continuously throughout the year, its energy consumption is about one-fifth of the US's annual power generation in 2025.

New AI computing power contracts add momentum, and the “AI+ Space” superempire's revenue target of 100 billion US dollars has entered a critical delivery stage

SpaceX Chief Financial Officer Brett Johnson said that the company is more confident that it can achieve an annual revenue operating rate of 100 billion US dollars, highlighting the strong bullish expectations of the company's executives and the market for the rapid growth of Elon Musk's supertech giant covering AI applications, huge AI computing resources, and commercial space and satellite internet.

Johnson said at the Goldman Sachs Communacopia+ technology conference that the company has signed a new agreement to provide unidentified customers with AI computing power, with a monthly amount of 1.11 billion US dollars starting in December. The contract corresponds to an annualized revenue of about 13 billion US dollars. He added at the conference that the recently signed agreement makes SpaceX “more confident” about achieving an annual operating rate of 100 billion US dollars in annual revenue.

He also discussed SpaceX's plans to expand terrestrial AI computing power capacity, saying the company expects to deploy slightly more than 2 gigawatts of capacity by the end of 2026, and set a capacity target of 5 to 10 gigawatts for 2027.

Johnson said that electricity supply is increasingly becoming a limiting factor for terrestrial AI infrastructure, which is one of the reasons SpaceX sees orbital computing as a long-term opportunity and explores it.

In terms of numbers, the annualized revenue target here is the annual operating scale calculated by multiplying monthly revenue by 12, which does not mean that the 2026 earnings report revenue will reach 100 billion US dollars. SpaceX may even need to generate around $8.3 billion in monthly revenue to reach this level. The AI, commercial space, and satellite operator reported second-quarter revenue of 7.8 billion US dollars, which is equivalent to about 2.6 billion US dollars per month, which means that monthly revenue needs to be increased to at least three times the original.

Existing agreements include a $6.7 billion AI cloud computing resource service agreement that will gradually accelerate in October, and an AI hosting agreement with a monthly amount of 1.11 billion US dollars starting in December. The company's huge AI computing power agreement with Google, the global leader in AI applications, and Anthropic, which is about to enter the stock market, has already brought in more than 2 billion US dollars in monthly revenue.

Whether the figure of 100 billion US dollars can be reached depends on how quickly the business scale of these agreements expands; delays in delivery or slowing growth may prevent SpaceX from reaching this level of annualized revenue.

Johnson previously said that since the company has signed a computing power rental contract, SpaceX's AI computing power infrastructure payback period may even be less than a year.

Wall Street financial giant Morgan Stanley's latest key trend and judgment around the OpenAI Astra model, which Nvidia CEO Hwang In-hoon called the “AGI Era,” is that the AI model's capabilities have been greatly enhanced to make more workloads economically viable, thereby strengthening the supply constraints of AI computing power, data center power chains, carrier boards, and storage manufacturing. According to its scenario calculation, the power capacity corresponding to the computing power deployment of hyperscale cloud vendors will expand from about 35 gigawatts in 2025 to about 145 gigawatts in 2028, reaching about 4.1 times the original.

The Astra release boosted the market's expectations for general artificial intelligence (AGI), leading to a stronger computing power demand trajectory. In particular, the new “pay-for-performance” growth model is expected to bring stronger demand for total computing power, while recent more direct evidence of computing power demand comes from the AI R&D process itself — that is, the “recursive self-improvement (RSI)” development trajectory where AI began to “create AI.”

As the advanced and cutting-edge model led by Astra brings more and more strong demand for AI computing power, Morgan Stanley expects the data center comprehensive capital expenditure of the four largest supercloud computing and AI application vendors in North America to rise from US$917 billion in 2026 to US$1.47 trillion in 2027 and US$1.64 trillion in 2028. The deployment capacity is expected to expand from 35 gigawatts in 2025 to 145 gigawatts in 2028.

The key to the sustainability of AI investment is whether return on capital and operating cash flow can support expansion. According to Damo scenario estimates, the model company uses its own infrastructure to provide API services, and the return on investment (ROIC) is about 46%, which is higher than 31% of cloud vendors renting GPUs and 25% of relying on third-party infrastructure to provide API services. This allows companies that master the model, computing power, and commercial channels at the same time to obtain higher returns, but these are quantitative model estimates. Meanwhile, in terms of the combined operating cash flow of the four giants, Damo is expected to increase from US$739 billion in 2026 to US$1.23 trillion in 2028, and demand for additional debt financing will drop from US$238 billion to US$90 billion.

Terrestrial computing power is beginning to be realized as AI orders grow, and the space vision expands the boundaries of future growth

SpaceX's ground-based AI computing power expansion plan provides Wall Street financial giants such as Morgan Stanley and Goldman Sachs with a specific measure to actively observe the needs of the industrial chain. SpaceX expects to deploy slightly more than 2 gigawatts of AI infrastructure by the end of 2026, with a target of 5 to 10 gigawatts in 2027.

A gigawatt measures the power scale of the relevant facility, and actual computing power also depends on chip configuration, cluster efficiency, and utilization. Judging from engineering and business logic, intelligent entities undertake more complex tasks of continuous operation and repeated use of tools, which will increase computing, storage, and network requirements; leading customers are purchasing capacity in advance, which is a positive sign that this demand has turned into commercial commitment. Corresponding construction can drive the procurement of AI core computing accelerators (GPU/TPU/XPU), data center high-performance HBM/DRAM, server memory, enterprise-grade SSD, optical interconnection, and power supply and distribution equipment, but the supplier's actual revenue still depends on the project's commissioning schedule and order share.

Space orbital AI computing resources are the long-term direction of SpaceX's attempt to break through terrestrial power constraints. The company has submitted an application to deploy up to 1 million orbital data center satellites. Recent reports mentioned that the first batch of related satellites is scheduled to be launched at the end of 2027. This is still a long-term or long-term vision-level construction plan.

The potential advantage of space AI data centers is to use suitable orbital sunshine conditions to reduce dependence on terrestrial power grid access and land resources; engineering difficulties focus on launch costs, power supply system quality, radiation environment, communication capabilities, and equipment renewal and maintenance. However, investors especially need an accurate understanding of heat dissipation: there is no air convection in a vacuum. Chip waste heat must be transferred to the radiator through a heat transfer system and then discharged by means of thermal radiation. The heat dissipation area and quality itself are limited by scale.

The underlying logic of the space AI data center is to obtain solar energy in orbit and complete the calculation locally, and send the data back to Earth to reduce dependence on terrestrial power grid expansion, land, and cooling water.

Choosing an appropriate sun-synchronous orbit in the morning and sunset can obtain close to continuous sunlight and reduce energy storage requirements; however, low temperatures in space do not mean that the chip can automatically cool: the electricity consumed by computing equipment is almost eventually converted into heat. The vacuum environment lacks air convection, and waste heat must be transmitted to the radiator through a heat pipe or fluid circuit, and then discharged as thermal radiation.

Therefore, the weight and deployment of large-scale heat dissipation structures, launch costs, chip radiation resistance, high-bandwidth interstellar communication, and in-orbit maintenance are still key bottlenecks in large-scale construction. Lightweight solar cells have undoubtedly greatly improved the power supply process, and the economy of the entire system still depends on the joint maturity of these engineering conditions.

Just after Nvidia, the “superpower of AI chips,” announced a strong performance report that continues to raise global AI capital expenditure expectations and a new round of AI computing power industry chain bullish market, Musk tried to push the boundaries of AI infrastructure expansion from the ground to space orbit — SpaceX (SPCX.US), which he founded and steered, plans to launch the first batch of AI data center satellites using Nvidia's next-generation computing power clusters — the Vera Rubin architecture AI GPU-led clusters, and will form a “significant scale” in 2028. This does not mean that terrestrial data centers will be quickly replaced, but rather that space orbital computing power can bypass major bottlenecks such as terrestrial power grids, land, and water, and become an additional supply layer for AI computing power.

The bottom judgment of SpaceX management, such as Musk, is that global data center computing power requirements may reach 235 gigawatts in 2030, 70% of which will be used for AI, while the Earth's power grid, land, approval, and environmental capacity are difficult to support terawatt expansion; the Sun accounts for about 99.8% of the solar system's energy. As a result, SpaceX plans to begin commercializing modular orbital computing power at the end of this decade. The long-term goal is to deploy 100 gigawatts of AI computing capacity to orbit every year — if it operates continuously throughout the year, its energy consumption is about one-fifth of the US's annual power generation in 2025.

The SpaceX prospectus clearly defines “building a space civilization that continues to expand, and eventually moving towards a Type II civilization that can use all of the Sun's energy” as a long-term paradigm shift; Musk himself has stated that the lunar satellite factory, Mass Driver (Mass Driver), and AI hardware deployment of more than 100 terawatts per year will push humans to achieve “substantial positive progress” towards Type II civilization.