June 5, 2026, marked the disclosure of one of the largest computing power collaborations in the global tech and aerospace sectors to date: Google and SpaceX officially signed a cloud computing service agreement set to last approximately three years. Under this deal, from October 2026 through June 2029, Google will pay SpaceX $920 million per month in exchange for access to roughly 110,000 Nvidia GPUs, along with supporting CPUs, memory, and other infrastructure. The total contract value over its lifecycle is about $30 billion.
The timing of this agreement is no coincidence. On June 12, SpaceX is set to debut on Nasdaq under the ticker SPCX at a fixed price of $135 per share, aiming to raise $75 billion and reach a valuation of around $1.77 trillion—making it the largest IPO in US stock market history. Securing Google as a major client right before this milestone carries dual strategic significance: it strengthens SpaceX’s valuation narrative and supports Google’s AI infrastructure ambitions.
This article explores how Starlink’s low-Earth orbit satellite network can fill coverage gaps left by traditional cloud providers, how locking in a major client pre-IPO boosts SpaceX’s valuation credibility, and how Google is extending its AI compute strategy beyond terrestrial data centers. We’ll also examine the potential impact of this deal on GOOGL and SPCX share prices, and outline how Gate’s stock trading features can help capture related asset volatility.
Transaction Overview: Key Terms and Data
Core Terms
According to regulatory filings submitted by SpaceX to the SEC, the agreement was officially signed on June 5, 2026. Key provisions include:
| Term | Details |
|---|---|
| Effective Date | Full monthly fees begin October 2026; ramp-up period before then with lower fees |
| Contract Duration | Until June 2029 |
| Monthly Fee | $920 million/month (from October) |
| Total Contract Value | Approximately $30 billion (32 months of full payment) |
| Compute Resources | About 110,000 Nvidia GPUs + CPUs, memory, and related components |
| Termination Clause | After December 31, 2026, either party may terminate with 90 days’ notice |
| Delivery Guarantee | If SpaceX fails to deliver the promised GPUs by September 30, Google may terminate or reduce payment after a one-month grace period |
| IP Ownership | Google retains full ownership of its content, AI models, and related data |
Second Major Compute Leasing Contract
Notably, this is SpaceX’s second major compute leasing deal in a short span. In May 2026, AI company Anthropic reached an agreement to lease all compute output from SpaceX’s Colossus data center cluster—previously built and operated by xAI—for $1.25 billion per month. Together, these contracts bring SpaceX about $2.17 billion in monthly compute leasing revenue, with an annualized recurring revenue (ARR) of roughly $26 billion. If both contracts run their full course, their combined value will exceed $70 billion.
Dimension One: Starlink’s Edge Node Coverage Complements Traditional Cloud Gaps
The Core Question: Why Doesn’t Google Just Build More Data Centers?
Google is hardly short on compute resources. Industry estimates suggest that, thanks to its ongoing TPU chip advancements, Google is among the world’s largest holders of AI compute power. In 2026, Google committed $175–185 billion in capital expenditures for global data center expansion, and Alphabet recently announced an $80 billion equity financing plan to support this investment.
However, AI compute demand continues to outpace industry expectations. A Google Cloud spokesperson described the deal as "a short-term, timely agreement to ensure transitional capacity for surging demand on our AI agent platform, Gemini Enterprise—a demand that has exceeded even our projections." Google Cloud’s latest earnings report shows its backlog (signed but not yet recognized contract value) nearly doubled quarter-over-quarter, now exceeding $460 billion.
The core issue is that terrestrial data centers are constrained by energy supply, land availability, cooling, and power infrastructure. As AI model parameters scale up, traditional data center expansion can’t keep pace with demand, both in speed and marginal cost.
Starlink’s Unique Asset: Coverage Where Cloud Providers Can’t Reach
Starlink now operates over 9,600 low-Earth orbit satellites, serving 10.3 million users across 160+ countries and regions. In 2025, this segment generated $11.4 billion in revenue and $4.4 billion in operating profit. Yet Starlink’s strategic value extends far beyond broadband access.
Traditional cloud providers (AWS, Azure, Google Cloud) cluster data centers in densely populated, well-powered, and well-connected areas. Many regions worldwide—including remote mines, maritime routes, polar research stations, and desert oil fields—lack the economic or technical feasibility to connect to these centers. In such scenarios, AI-driven edge computing is becoming a real need.
Starlink’s satellite network offers global coverage by design. Paired with Google’s AI inference capabilities, the two can deploy edge AI workloads on Starlink satellites or ground terminals—for example, real-time analysis of satellite imagery in remote areas, AI-assisted navigation on ships, or maintaining emergency AI response in disaster zones with disrupted communications. This edge AI architecture eliminates the need to send raw data thousands of miles back to terrestrial data centers, enabling effective AI model operation even in low-bandwidth, high-latency environments.
From Edge to Orbit: The Long-Term Vision for Space-Based Compute
While this deal is fundamentally about compute leasing, it points to a shared long-term vision for Google and SpaceX: orbital data centers. In May 2026, the two began preliminary talks about building data centers in orbit, leveraging the Starlink network and low-Earth orbit platforms to address the energy, location, and bandwidth limitations of terrestrial centers.
The core idea is to deploy AI compute in low-Earth orbit, powered by near-continuous solar energy and connected by high-speed inter-satellite laser links (ISL). Starlink’s current laser terminals support 25 Gbps per channel. This aligns closely with Google’s "Suncatcher" orbital data center project, which aims to launch a constellation of solar-powered satellite data centers by 2027.
Market research estimates the global orbital data center market at $1.28 billion in 2025, growing to $3.81 billion by 2034 (CAGR: 12.96%). The broader market for space-based edge computing is even larger, reaching $168.91 billion in 2025 and projected to hit $345.04 billion by 2034. This expansive market provides a long-term growth narrative for the Google-SpaceX partnership.
Dimension Two: SpaceX’s Pre-IPO Client Win and Valuation Credibility
SpaceX’s Valuation: A High-Stakes Pricing Debate
SpaceX plans to issue 555.6 million shares at $135 each, targeting a $75 billion raise and a $1.77 trillion valuation. This valuation is hotly debated:
- Bullish Case: SpaceX’s S-1 filing claims a total addressable market (TAM) of $28.5 trillion, spanning broadband, mobile communications, AI infrastructure, and enterprise applications.
- Bearish Case: Morningstar analysts use a discounted cash flow model to value SpaceX at about $780 billion—roughly 44% of the IPO price. Morningstar states, "We believe the company is significantly overvalued and expect investors will have the chance to buy in at more attractive prices post-IPO."
The financials explain the controversy. SpaceX posted a net loss of $4.28 billion last quarter, is projected to lose $4.94 billion in 2025, and has accumulated $41.3 billion in deficits. Its AI division, xAI, lost about $2.5 billion in a single quarter. With 2025 revenue at $18.7 billion, the $1.77 trillion valuation implies a price-to-sales ratio of about 92x—extremely high, even for a growth company, and particularly risky for a conglomerate still running large losses.
Compute Leasing Contracts Bolster Valuation Narrative
Against this backdrop, the Google and Anthropic contracts—together worth $26 billion in annualized revenue—offer two key supports for SpaceX’s valuation:
First, revenue visibility. According to SpaceX’s IPO filing, its business is divided into three segments: Starlink ($11.4 billion in 2025 revenue), space launch (over $4 billion), and AI/xAI (about $3.2 billion). The annualized revenue from these two compute contracts ($26 billion) already surpasses the combined revenue of its other segments. This means that, even if rocket and satellite growth stalls, compute leasing alone could underpin SpaceX’s valuation base.
Second, a strategic business model shift. Traditionally, SpaceX was seen as a "rocket company + satellite broadband operator"—a positioning that can’t justify a valuation premium over Nvidia. The compute leasing deals show SpaceX repositioning itself as an "AI infrastructure-as-a-service" provider. By converting xAI’s Colossus data center cluster (originally built for Grok model training) into a third-party compute leasing asset, SpaceX shifts its data center infrastructure from a cost center to a profit center.
This shift is profound: SpaceX’s AI infrastructure is no longer just for internal model development but is now monetized at scale with clients like Google. Google Cloud’s spokesperson described the deal as "bridging capacity," hinting at a timing gap between Google’s own data center expansion and surging AI demand—precisely when SpaceX can provide mature compute resources.
Issuer Risk Disclosure: Contract Termination Flexibility
It’s important to note that these contracts are not irrevocable long-term commitments. The agreements specify that after December 31, 2026, either party can terminate with 90 days’ notice. If SpaceX fails to deliver the required GPUs by September 30, Google can terminate or reduce payments after a one-month grace period. This means contract enforceability beyond 2027 is uncertain. Investors should factor in these termination options when modeling the full $70+ billion contract value.
Dimension Three: Google’s AI Compute Strategy Beyond Data Centers
The "Asymmetric" Compute Supply Landscape
Currently, AWS holds about 30% of the global cloud market, Azure 25%, and Google Cloud 13%. While Google Cloud trails in market share, its custom TPUs and Gemini model ecosystem give it unique AI compute strengths. At the 2026 Google Cloud Next conference, Google unveiled the 8th-gen TPU—dual-chip architecture optimized for training (TPU 8t) and inference (TPU 8i)—and, for the first time, announced direct TPU hardware sales to select clients (previously only available via cloud rental).
However, AI compute supply is entering an "asymmetric competition" phase. Traditional cloud providers’ advantages are built on data center scale, but data centers face physical limits: land, power, cooling, and carbon emissions. Google’s "Project Suncatcher"—an orbital compute constellation powered entirely by solar energy—represents a long-term bet to break these constraints. If successful, orbital data centers will no longer be limited by grid capacity, land supply, or cooling water, enabling fundamental differentiation in the compute supply chain.
Starlink as the "First Mover" in Space-Based Compute
Before "Project Suncatcher" is fully realized, Starlink already provides a ready-made orbital infrastructure.
Architecturally, Starlink’s low-Earth orbit network is more than just communications. Its satellites feature inter-satellite laser links, some onboard compute, and high-precision attitude control—all foundational for future orbital data centers. SpaceX CEO Elon Musk has publicly stated plans to expand the Starlink V3 fleet and develop orbital data centers to meet surging global AI compute demand.
Another proof point is Starcloud (formerly Lumen Orbit), which has ordered over 50 Starlink mini laser terminals from SpaceX to build its own orbital data center network, with the first hardware set to launch within a year. This demonstrates that Starlink’s inter-satellite communications infrastructure is technically ready to support third-party orbital data center deployments.
For Google, piloting edge AI workloads on Starlink satellites offers hands-on experience, technical data, and business model validation ahead of large-scale "Project Suncatcher" deployment. This "test first, scale up" approach is more prudent technically and more feasible commercially.
GOOGL and SPCX: Dual Impact on Share Prices
Impact on GOOGL (Alphabet)
Short term: Modest positive financial impact. The $920 million monthly compute spend is less than 0.3% of Alphabet’s projected 2025 revenue ($350+ billion), so the immediate financial effect is limited. Strategically, however, the deal alleviates Gemini Enterprise’s compute bottleneck. Rapid growth in Google Cloud’s backlog shows real AI service demand is accelerating, and ample compute supply helps convert backlog into recognized revenue—directly supporting Google Cloud’s growth.
Medium term: Shifting strategic cost structure. Google’s choice to procure externally rather than build everything in-house reflects the growing industry norm of a "build + buy" hybrid AI infrastructure model. If Google can source compute faster and at lower marginal cost externally, it boosts capital efficiency; conversely, over-reliance on external suppliers like SpaceX could affect long-term control and cost structure.
Risk note: The 90-day termination clause gives Google flexibility. If its own data center build-out outpaces expectations, or if in-house TPUs prove more cost-effective, Google may scale back or end the contract after 2027. For now, Google frames this as "bridging capacity"—a transitional supplement, not a long-term dependency.
Impact on SPCX (SpaceX)
Strengthening IPO pricing logic. As discussed, the two compute leasing contracts add about $26 billion in annualized revenue—a key rationale for launching the IPO at a 92x price-to-sales ratio. Without this income, SpaceX’s annualized revenue would be $18.7 billion, pushing the valuation multiple even higher. Thus, the contract is critical to justifying the IPO price.
Sustainability challenges. Morningstar values SpaceX’s core businesses (launch + Starlink) at $611 billion and AI at $170 billion (probability-weighted), for a total of $781 billion. The roughly $1 trillion gap with the IPO valuation must be bridged by delivering on AI business performance—highly dependent on the actual execution of the Google and Anthropic contracts and SpaceX’s ability to win more third-party compute clients.
Peer comparison: Nvidia is currently valued at about $5.3 trillion with $250 billion in annual revenue (21x price-to-sales); Anthropic’s latest funding round valued it at $965 billion. SpaceX’s 92x multiple far exceeds these, reflecting a premium for "continued Starlink growth + scaled AI profitability + orbital data center commercialization"—all expected to play out between 2027 and 2029.
Gate Platform Stock Trading Features and How to Trade
For users looking to capitalize on GOOGL and SPCX investment opportunities, Gate offers an integrated stock trading solution.
Core Advantages of Gate Stock Trading
- Unified Multi-Asset Trading: Seamlessly switch between crypto and US stocks in a single account—no cross-platform transfers required for dual exposure to digital and traditional assets.
- 24/7 Trading Window: Unlike traditional brokers limited to US market hours (9:30 p.m.–4:00 a.m. Beijing time), Gate’s stock trading feature offers round-the-clock order management.
- Fractional Trading: Trade in increments as small as $1, enabling diversified investments even in high-priced stocks (GOOGL at ~$200/share, SpaceX IPO at $135/share).
- Real-Time Quotes and Depth: Access live prices, pre-market and after-hours data, company announcements, and integrated analyst reports—so you’re always up-to-date on events like those discussed here.
- Global Market Coverage: In addition to US stocks, trade Hong Kong and other major markets, supporting your global asset allocation needs.
GOOGL and SPCX Trading Reference Guide
- Account Setup: Complete Gate account registration and KYC, then deposit USD or USDT as trading margin.
- GOOGL Trading Path: Search for "GOOGL" to access the trading page; choose market or limit orders. Consider Google Cloud’s quarterly results and Gemini Enterprise subscription data when timing entries.
- SPCX Trading Path: SpaceX is set to list on Nasdaq as "SPCX" on June 12; find "SPCX" under Gate’s stocks section to participate in this historic IPO on day one.
- Risk Management Advice: Stock trading carries inherent market risks. Users should tailor position management strategies to their own risk tolerance and avoid over-concentration in a single asset.
Conclusion
The $30 billion compute contract between Google and SpaceX represents a strategic alignment between two giants with deeply complementary resources, set against the backdrop of the "post-Moore’s Law" compute bottleneck.
Starlink’s low-Earth orbit network provides physical infrastructure for edge AI in regions beyond the reach of traditional cloud providers, addressing structural coverage gaps. Google’s AI model ecosystem and Gemini application demand offer SpaceX’s data center assets a sticky commercialization channel. Over the long term, this deal lays both the technical and commercial groundwork for the partners’ shared vision of orbital data centers.
For investors, GOOGL offers stable exposure to the AI application layer, while SPCX represents a high-upside option on the space-based infrastructure frontier. When balancing between the two, it’s essential to consider macro factors like AI compute demand growth, data center expansion efficiency, and regulatory policy. Gate’s stock trading features provide the technical tools and flexible, fractional trading needed to implement such a strategy with ease.
