Sequoia Partner: SpaceX Stands on the Eve of an NVIDIA-Style Growth Explosion, With Starship, AI Computing Power, and Direct-to-Cell Satellite Connectivity Severely Undervalued

Written by Zhao Ying, Wallstreetcn

Recently, Sequoia Capital partner Shaun Maguire posted on X that SpaceX is on the eve of a revenue explosion comparable to NVIDIA's explosive growth over the past three years, while most investors severely underestimate this potential.

Following SpaceX's earnings call, Maguire publicly released a bottom-up financial model, claiming the figures it presents are "breathtaking." He noted that the market is gripped by panic due to a series of macro factors including geopolitical uncertainty and inflationary pressures, yet he chose to speak against the current trend at this moment, releasing predictions that he believes will seem "crazy to more than 90% of investors."

Maguire specifically highlighted three core theses that he believes are the most underestimated by the current market: top-tier talent density, SpaceX's position as a low-cost supplier in the AI computing infrastructure sector, the historic moat built by Starship, and the unit economics of Starlink. He projects that the Direct-to-Cell business will surpass Starlink's revenue scale in virtually any scenario between 2029 and 2030.

Top-Tier Talent Density: SpaceX's Implicit Moat

Maguire characterizes SpaceX's team as "the most elite team on Earth," distinctly separating its intensity-driven culture from that of other tech companies in Silicon Valley.

He argues that what sets SpaceX apart is its combination of extreme work intensity with an ultimate objective evaluation mechanism—rockets either fly or explode, leaving no room for political maneuvering. This hardwired physical feedback mechanism prevents mediocre performers from surviving internally, thereby cultivating an exceptionally high density of talent over 25 years of accumulation.

In the wave of large-scale AI computing infrastructure construction, Maguire considers this factor to be decisive. While Amazon, Google, and various cloud providers possess some outstanding talent, at SpaceX "everyone is excellent." Faced with the construction challenges of gigawatt-scale AI data centers—which he believes represents an increase of one to two orders of magnitude over past megawatt-scale CPU data centers—SpaceX's ability to handle problems at the "atomic level" leaves other competitors far behind. Shortages of turbines, cross-state power transmission, transnational shipping of large equipment... in his view, the SpaceX team has solutions for all these challenges.

Starship: The Largest Technological Moat in Human History

In his article, Maguire devotes substantial space to arguing that Starship is the most complex system humanity has ever built, surpassing widely recognized technological pinnacles such as cutting-edge chip manufacturing, ASML's EUV lithography machines, and ITER.

His evaluation framework asks: If a group of the world's finest talents were given unlimited resources to catch up to a specific technological frontier, how long would it take? In front of Starship, he believes the answer exceeds that of any other technology sector.

Maguire cites the global competitive landscape as corroborating evidence: SpaceX achieved its first booster recovery in 2015, with competitors taking approximately a decade later to reach the same milestone. He believes that as Starship continues to advance, this gap is widening rather than narrowing.

He emphasizes that Starship's supply chain is highly vertically integrated within SpaceX, making the depth of the moat "virtually unimaginable."

Low-Cost Supplier Position: Structural Advantages in the Computing Market

Maguire believes that becoming a low-cost supplier in the commodities market is a highly strategic position. He analogizes the pricing dynamics of AI computing to oil production rather than broadband during the dot-com bubble era—arguing that the entry barriers for gigawatt-scale AI computing are far higher, the supply chain is tighter, and the construction cycle is longer.

His judgment is that even if computing costs drop to approximately $10 billion per gigawatt, SpaceX, as the low-cost and most flexible participant, will retain a competitive advantage. He anticipates that for the next four years, computing costs are unlikely to fall below $20 billion per gigawatt, and within this price range, he believes SpaceX's capital expenditure per unit will decrease further as competition subsides.

He constructs a company-wide forecasting model extending to 2028, noting that revenue per gigawatt and capital expenditure per gigawatt form a pair of "coupled constants"—if the former decreases, the latter will decrease proportionally, while the total gigawatts deployed by SpaceX will rise, partially offsetting the revenue loss.

Starlink Unit Economics: A Single Launch Leveraging Billions in Revenue

Maguire disclosed a set of core metrics to illustrate Starlink's capital efficiency. The current Starlink constellation generates approximately $20 billion in annual revenue, built through roughly 400 Falcon 9 launches.

Since a single Starship launch has a payload capacity roughly equivalent to 20 Falcon 9 launches, theoretically only about 20 Starship launches would be needed to sustain the $20 billion annual revenue. This implies roughly $1 billion in annual revenue per launch, which translates to $5 billion in lifetime revenue per launch when amortized over a five-year lifespan.

He states that as Starship enters commercial operation, costs for Starlink, Direct-to-Cell, and orbital computing will plummet significantly, expanding the addressable market accordingly. In his scenario forecasts, Direct-to-Cell revenue will surpass Starlink's core business by 2029 to 2030.

Maguire also points out that the Orbital Compute business will begin to feature prominently in financial models starting in 2029, and he projects that over the next decade, SpaceX's degree of vertical integration in the chip sector will match its current status in the rocket industry.

Join the official Coincamps community:

X: https://x.com/coincamps

Telegram: https://t.me/coin_camps