SpaceX (SPCX.US) CFO Says Confidence in $100 Billion Annual Revenue Target Grows! Ground Computing Power Secures Large Orders, Expanding Frontier for Space AI Vision

According to zhitongcaijing.com, the latest developments disclosed by SpaceX (SPCX.US)—founded and led by Elon Musk and a leader in AI and space exploration—at the highly anticipated Goldman Sachs Communacopia+ Technology Conference, show that a newly signed AI computing infrastructure supply agreement will generate $1.11 billion in monthly revenue starting in December, which annualized to $13.32 billion for a full 12 months. This has bolstered management’s confidence in achieving a $100 billion annual revenue run rate and highlights how AI computing infrastructure-related business is rapidly becoming the company’s core growth engine. SpaceX CFO Bret Johnson also announced key schedules for major growth business lines including ground computing capacity expansion, orbital AI deployment, large-scale commercialization of Starship, and satellite-to-mobile connectivity.

The $100 billion target is SpaceX’s goal for annualized revenue run rate by the end of 2026. The key to this strong growth target is revenue growth rate: to hit this target, SpaceX will need about $8.33 billion in monthly revenue—approximately 3.2 times the company’s Q2 average monthly revenue of $2.6 billion.

Musk previously posted on X predicting that SpaceX’s annual revenue will reach about $3.5 trillion around 2033, a much more aggressive target than his earlier projection of “$1 trillion by 2030.” Using Wall Street’s estimate of $100 billion in 2027 revenue as a base, this implies a 35x increase over six years, with an 81% compound annual growth rate. However, this aggressive forecast is Musk’s personal prediction and not official company guidance.

The Google and Anthropic compute agreements correspond to around $920 million and $1.25 billion per month, respectively. If the new contract starts as scheduled and the first two businesses maintain their scales, the three combined would be about $3.28 billion per month. These numbers show that external customer procurement is becoming a major support for SpaceX’s revenue growth. However, the $6.7 billion AI cloud service agreement does not have a clearly defined contract term in currently available SpaceX disclosures, so it cannot be directly counted toward monthly or annual revenue estimates.

These positive developments support the extension of SpaceX’s growth sources further into commercial compute services while leaving long-term upside potential for orbital computing—in other words, ground computing contracts support near-term revenue expansion, and the orbital AI data center vision actively explores long-term growth. For AI compute chain investment cases, SpaceX’s large AI cloud procurement deals support a markedly optimistic outlook for the AI computing industry chain.

At the Goldman Sachs Communacopia+ Technology Conference, SpaceX demonstrated strong confidence in reaching its $100 billion annual revenue run rate, and the company also plans to expand its ground AI infrastructure capacity from just above 2 GW at the end of 2026 to between 5 and 10 GW in 2027, and is exploring orbital computing to break ground power constraints. SpaceX plans to start commercialization of modular orbital compute at the end of this decade, with a long-term goal to deploy 100 GW of AI compute to orbit each year—if operated continuously, this would consume about one-fifth of the total U.S. annual electricity generation in 2025.

New AI compute contracts provide more confidence as the “AI + Space” super-empire’s $100 billion revenue goal enters critical delivery phase

SpaceX CFO Bret Johnson stated the company is even more convinced it can achieve the $100 billion annual revenue run rate target, highlighting both management’s and the market’s strong bullish expectations for the rapid growth of this AI application, massive AI compute resource, commercial space and satellite internet tech giant under Elon Musk.

At the Goldman Sachs Communacopia+ Technology Conference, Johnson said the company had signed a new AI compute supply agreement with an undisclosed customer, valued at $1.11 billion per month starting in December. The contract implies annualized revenue of about $13 billion. He added that this latest deal makes SpaceX “even more confident” in hitting the $100 billion annual run rate target.

He also discussed plans to further expand SpaceX’s ground AI compute capacity, stating that the company aims to deploy just over 2 GW by the end of 2026, with a 2027 goal of 5 to 10 GW.

Johnson stated that power supply is increasingly becoming a limiting factor for ground AI infrastructure, which is one reason why SpaceX sees orbital computing as a long-term opportunity and is actively exploring it.

Numerically, the annual run rate target is calculated by multiplying monthly revenue by 12; it does not mean reported revenue in 2026 will reach $100 billion. SpaceX may even need to generate around $8.3 billion monthly to reach this level. The company’s Q2 revenue was $7.8 billion—about $2.6 billion per month—meaning monthly revenue will need to more than triple.

Existing agreements include a $6.7 billion AI cloud resource service agreement ramping up from October, and an AI hosting contract worth $1.11 billion per month from December. Mega AI compute deals with global leaders Google and the soon-to-list Anthropic are already delivering over $2 billion in monthly income.

Whether the $100 billion mark can be reached depends on the scale-up speed of these agreements; delays or slowdowns in growth could prevent SpaceX from reaching this annualized revenue level.

Johnson previously said that because the company has already signed compute leasing agreements, the payback period for SpaceX’s AI compute infrastructure investments could be less than a year.

Wall Street giant Morgan Stanley’s latest projections in relation to OpenAI’s Astra large model—which Nvidia CEO Jensen Huang has called the beginning of the “AGI era”—is that major advances in AI models are making new workloads economically viable, strengthening the supply constraints of AI compute, data center power chains, substrate and memory manufacturing. In its scenario analysis, hyperscale cloud providers’ compute deployments could drive total data center power from about 35 GW in 2025 to approximately 145 GW in 2028—a 4.1x increase.

Astra’s launch boosted market expectations for Artificial General Intelligence (AGI), resulting in an even stronger compute demand trajectory, especially as the new “results-based billing” model could lead to even greater overall demand. The most direct recent evidence of compute demand comes from the AI development process itself—namely, the “recursive self-improvement (RSI)” R&D trajectory where AI is now “building AI.”

As the most advanced models like Astra drive surging AI compute demand, Morgan Stanley expects total data center capital expenditures by North America’s four super cloud and AI firms to rise from $917 billion in 2026 to $1.47 trillion in 2027 and $1.64 trillion in 2028, with deployed capacity rising from 35 GW in 2025 to 145 GW in 2028.

The sustainability of AI investment hinges on whether returns on capital and operating cash flow can support expansion. Morgan Stanley scenario analysis shows that companies using their own infrastructure to provide API services can reach a return on invested capital (ROIC) of about 46%, higher than the 31% for cloud providers renting out GPUs and the 25% for those depending on third-party providers. This supports higher returns for those who own the model, the compute, and commercialization channels—but these are model-based estimates. Meanwhile, the combined operating cash flow of the four giants is projected to increase from $739 billion in 2026 to $1.23 trillion in 2028, while additional debt financing needs fall from $238 billion to $90 billion over the same period.

Ground compute is now translating into AI order growth, while the orbital vision broadens the future growth boundary

SpaceX’s ground AI compute expansion plans offer detailed metrics for industry chain demand monitoring for Wall Street giants like Morgan Stanley and Goldman Sachs. SpaceX expects to deploy just over 2 GW of AI infrastructure by the end of 2026, targeting 5 to 10 GW in 2027.

Gigawatt measures power capacity for such facilities, but real compute power also depends on chip configurations, cluster efficiency, and utilization. From an engineering and business perspective, agents performing increasingly complex, persistent tasks will drive up compute, storage, and network requirements; large clients pre-booking capacity is a strong signal of this demand turning into committed business. Construction corresponds to procurement of AI compute accelerators (GPU/TPU/XPU), high-performance HBM/DRAM, server memory, enterprise SSDs, optical interconnects, and power distribution equipment, but suppliers’ actual profits depend on project commissioning schedules and order shares.

Orbital AI compute resources are SpaceX's long-term direction for breaking ground power constraints. The company has applied to deploy up to a million orbital data center satellites, with reports of the first batch set for launch by end of 2027—still a long-term construction plan at this stage.

The potential advantage of space-based AI data centers is leveraging favorable solar conditions in orbit, reducing reliance on terrestrial grids and land. Engineering challenges focus on launch costs, power supply quality, radiation environments, communication capabilities, and system maintenance and upgrades. But investors must understand thermal management: in vacuum, there is no air convection, so chip waste heat must be conducted to radiators and emitted as infrared radiation—the radiator’s size and mass are a limiting factor for scale.

The underlying logic of space-based AI data centers is to collect solar energy in orbit, carry out computing locally, then send data back to Earth, thereby reducing demand for land, grid capacity, and cooling water expansions on the ground.

Choosing suitable sun-synchronous dawn-dusk orbits can achieve near-continuous sunlight and reduce energy storage requirements; but cold temperatures in space don’t mean chips can cool themselves: almost all electric energy consumed by computing devices converts to heat, and with no air convection in vacuum, waste heat must move via heat pipes or fluid loops to radiators to be discharged as thermal radiation.

Therefore, the mass and deployment cost of large radiators, chip radiation resistance, high-bandwidth inter-satellite communications, and in-orbit servicing remain key bottlenecks for scaling. Lightweight solar panels have greatly improved power supply, but the system’s economics remain contingent on overall engineering advances.

Just as “AI chip superpower” Nvidia released strong results and raised global AI capex expectations—fueling the next AI compute industry bull market—Musk is now pushing AI infrastructure expansion from ground to orbit. His SpaceX (SPCX.US) plans to launch the first batch of AI data center satellites built using Nvidia’s next-gen Vera Rubin architecture AI GPU clusters in Q4 2027, reaching “significant scale” in 2028. This won’t instantly replace ground data centers, but bets that orbital compute can bypass key bottlenecks like grid, land, and water, creating a new supply layer for AI compute.

The fundamental view of Musk and SpaceX management is that by 2030, global data center compute demand could reach 235 GW, 70% of which for AI, while Earth’s grids, land, permitting, and environment can’t support terawatt-scale expansion; the sun supplies about 99.8% of our solar system’s energy. SpaceX thus plans to commercialize modular orbital computing by the end of the decade, with a long-term goal to deploy 100 GW of AI compute in orbit annually—the power demand would be about a fifth of total U.S. electricity generation in 2025 if run continuously.

The SpaceX prospectus explicitly frames “building an ever-expanding spacefaring civilization, ultimately advancing to a Type II civilization harnessing the sun’s full energy” as its vision for a paradigm shift; Musk has also said lunar factory satellites, mass drivers, and annual AI hardware deployments exceeding 100 terawatts will push humanity toward “meaningful positive progress” toward a Type II civilization.