스페이스X, AI 인프라 판매로 280억 달러 매출 확보…시장 반응은?
SpaceX Made a Huge Pivot in Its Business, but the Market Hasn't Caught on Yet
스페이스엑스가 AI 컴퓨팅 인프라를 판매하며 연간 280억 달러의 매출 계약을 확보하면서 단기적으로 매출 증대 가능성은 높으나 장기적인 비즈니스 모델에 대한 의문이 제기되고 있습니다.
핵심 요약
스페이스X는 AI 기업들과의 계약으로 연간 280억 달러 매출을 확보하며 인프라 판매로 비즈니스 모델을 전환했습니다.
핵심요약
- 연간 280억 달러 매출 확보 계약 체결
- AI 서비스가 아닌 컴퓨팅 인프라 판매로 전환
- 데이터센터에 사용되지 않는 GPU 수백만 대 보유
- 단기 매출 증가 가능성 vs 장기 비즈니스 가치 우려
도입
이번 기사에서 스페이스X의 AI 부문 전략 전환이 주목받고 있습니다. 특히, AI 기업들에게 컴퓨팅 인프라를 판매하는 계약으로 280억 달러의 매출을 확보한 점은 투자자에게 중요한 정보를 제공합니다. 이 전략이 스페이스X의 재무 상태와 주가 성장에 미치는 영향에 대해 심층적으로 분석해보겠습니다.
본문 1: AI 인프라 판매 계약의 단기적 영향
스페이스X는 최근 AI 기업들과의 계약으로 연간 280억 달러의 매출을 확보했습니다. 이는 기존의 AI 서비스를 제공하는 모델에서 벗어나, 컴퓨팅 인프라를 판매하는 '네오클라우드' 모델로의 전환을 의미합니다. 이 전환은 단기적으로 매출 증가에 기여할 수 있지만, 장기적인 경쟁력 확보에 대한 의문이 제기되고 있습니다. 특히, 경쟁사들과의 차별화된 기술력이나 서비스 제공 능력의 부재가 문제될 수 있습니다.
본문 2: 장기적 비즈니스 모델의 지속 가능성
스페이스X의 AI 부문의 전략 전환은 단기적인 매출 증가를 목표로 하고 있지만, 장기적인 비즈니스 모델의 지속 가능성에 대한 우려가 있습니다. 특히, 컴퓨팅 인프라 판매는 상대적으로 저차별화된 서비스이며, 경쟁사들과의 가격 경쟁력이 중요해질 수 있습니다. 또한, AI 기술의 빠른 발전으로 인해 인프라 수요가 변동할 가능성도 고려해야 합니다. 이는 스페이스X의 재무 안정성과 주가 성장에 부정적인 영향을 미칠 수 있습니다.
결론
스페이스X의 AI 부문 전략 전환은 단기적인 매출 증가를 가져올 수 있지만, 장기적인 비즈니스 모델의 지속 가능성에 대한 우려가 있습니다. 투자자들은 이 전략이 스페이스X의 재무 상태와 주가 성장에 미치는 영향을 주의 깊게 관찰해야 합니다. 특히, AI 기술의 발전과 시장 수요 변화에 대한 감시를 강화하는 것이 중요합니다.
Original Article
SpaceX Made a Huge Pivot in Its Business, but the Market Hasn't Caught on Yet
There's no shortage of excitement about the revenue potential for Space Exploration Technologies ( SPCX +0.13% ) , better known as SpaceX. Some analysts see it reaching trillions of dollars in annual sales across its launch services, connectivity business, and artificial intelligence segment.
That said, nobody expects it to be a straight line up and to the right. In fact, SpaceX could see a major revenue inflection this year thanks to a huge pivot in its AI segment.
The shift speaks volumes about the current state of the business and its potential going forward. And while the market is cheering some of the headline numbers, it may be underappreciating the implications for the company's finances and the stock's performance.
Image source: The Motley Fool.
The biggest revenue growth driver at SpaceX
Over the last few weeks, SpaceX has signed several deals to provide compute to other AI companies :
Those three contracts add up to nearly $28 billion in annual revenue, and SpaceX could look to make even more deals. "We have sufficient capacity to provide compute for our own AI models, including support of our training and inference demands, and to satisfy the obligations under these agreements," management wrote in its IPO registration filing with the SEC.
The shift is clear: The growth in SpaceX's AI revenue will stem from selling infrastructure rather than its own AI services. In effect, SpaceX is becoming another infrastructure-as-a-service business, or "neocloud," selling relatively undifferentiated compute capacity to leading AI labs that need as much as possible. Meanwhile, its own AI model is seeing so little use that it has hundreds of thousands of GPUs sitting idle in its data centers.
That shift will provide a significant boost to revenue right now, but investors need to consider how valuable that revenue and line of business are for SpaceX.
Does SpaceX have a competitive advantage?
The biggest concern for operating a neocloud company is establishing a competitive advantage. For the most part, the end product is undifferentiated. A token is a token. So, the only way to compete effectively is to keep your costs lower than everyone else.
To that end, SpaceX believes it has an advantage. "Our first-principles thinking enables us to build coherent compute at scale and at rapid speed with lower costs than most other companies in the industry," management wrote in its S-1 filing. However, SpaceX's strategy of retrofitting old factories near efficient water sources won't scale. There are only so many prime locations to construct data centers.
SpaceX may be able to find a sustainable cost advantage with orbital data centers. The cost will depend heavily on its ability to bring launch costs down by developing its fully reusable heavy-lift rocket, Starship. Management says it could launch its first orbital data centers by 2028.
What does it all mean for SpaceX stock?
The move to monetize its limited compute capacity by renting it out instead of using it for its own AI development is telling. Management expects its large language model Grok to remain a niche player among large language models. That severely limits its profit potential . With the majority of SpaceX's revenue growth stemming from its pivot toward becoming a neocloud operator, investors may want to reassess the value of that revenue growth.
Leading neocloud providers CoreWeave and Oracle have massive backlogs of compute contracts.
CoreWeave has contracted revenue of nearly $100 billion as of the end of the first quarter, with $36 billion of that to be recognized over the next two years. The total backlog is larger than SpaceX's, but its annual run rate is smaller. For reference, CoreWeave's market cap sits around $55 billion.
Oracle's remaining performance obligations total a whopping $638 billion as of the end of its most recent quarter. It's adding tens of billions of dollars in contracts every quarter. Approximately $75 billion of that backlog will be recognized over the next 12 months, and over $200 billion will hit Oracle's top line in the subsequent two years.
CoreWeave's and Oracle's strong revenue growth suggest SpaceX could achieve similar results over the coming years. However, that revenue isn't worth nearly as much as the revenue that comes from selling AI software services themselves.
That's evidenced by the price-to-sales multiples assigned to CoreWeave and Oracle: 4.2 and 5, respectively. By comparison, SpaceX trades at well over 100 times trailing 12-month sales. Even if it triples its sales over the next year thanks to infrastructure contracts, its sales multiple still sits well above reasonable levels.