Amazon has signed a deal with rival SpaceX to launch some of its planned 3,236 internet-beaming satellites, swallowing its pride to get its Project Kuiper constellation up and running.
The Deal
- Amazon will pay SpaceX to launch 47 Kuiper satellites on 3 Falcon 9 rockets
- Launches expected to take place between late 2023 and 2024 from Cape Canaveral
- Kuiper is Amazon’s $10+ billion project to provide global broadband internet coverage
- SpaceX seen as having most reliable large payload launch capacity
Faced with development delays for its own rockets, Amazon (AMZN) announced Friday it is contracting with Elon Musk’s SpaceX to carry some of its planned Project Kuiper internet satellites into orbit.
Amazon said it has signed a deal for "three Falcon 9 launches to support our initial satellite deployment." That will put 47 production design Kuiper satellites into orbit to start building out the full constellation of 3,236 envisioned satellites.
The three Falcon 9 rockets will launch from SpaceX’s Cape Canaveral facilities between late 2023 and 2024. The companies did not disclose financial terms.
Why SpaceX?
Industry analysts say Amazon really had little choice but to turn to the launch leader if it wants to get Project Kuiper off the ground.
Bezos’ Blue Origin is still far from ready to handle big satellite constellation launches. And United Launch Alliance — the Lockheed-Boeing joint venture providing Pentagon launches — has a large backlog.
That left SpaceX, which already has put up almost 3,000 Starlink internet satellites with Falcon 9, as only realistic choice, despite the bitter rivalry between space billionaire bosses Elon Musk and Jeff Bezos.
“SpaceX offers launch capabilities that does not seem to have alternatives at that scale,” said industry analyst Quilty Analytics’ Chris Quilty. “If [Project Kuiper] cannot launch at scale and with frequency, the business case likely falls down.”
Table: Launch Providers Capability Comparison
| Provider | Max Payload | Launches 2022 | Success Rate | Cost per launch | Notes |
|---|---|---|---|---|---|
| SpaceX Falcon 9 | 22,800 kg to LEO | 61 | 98% | $60-90 million | Workhorse vehicle, flies frequently |
| ULA Atlas V | 18,850 kg to LEO | 6 | 100% | $150-300 million | Reliable but more expensive, less capacity |
| Blue Origin New Glenn | 45,000 kg to LEO | 0 | Unproven | Unknown | Still in testing, not ready yet |
Quilty thinks Amazon may need 200-400 launches over five years — meaning it likely needs multiple launch providers. But SpaceX with its flight-proven reusable boosters has unrivaled launch cadence.
Amazon’s Project Kuiper chief said as much in a statement on the deal: “Securing launch capacity from multiple providers has been a key part of our strategy from day one,” said Kuiper’s senior vice president Rajeev Badyal. “[SpaceX] has proven themselves to be an industry leader.”
Industry Shakeup
Coming from arch tech rival Bezos, the SpaceX deal marks a humbling moment.
Bezos founded Blue Origin two decades ago with the explicit goal of supporting space ambitions like Kuiper. But its heavy-lift New Glenn rocket intended for satellite mega-constellations is seriously delayed.
The SpaceX deal essentially cedes a market from one billionaire’s space company to another’s. It also hands more leverage to the leading launch provider — and one with its own growing Starlink global internet satellite business, which SpaceX could prioritize over Amazon.
“We knew this day was coming, there would have to be coopetition between the version two constellations,” said industry consultant Caleb Henry of Quilty Analytics.
Race to High-Speed Internet
The background is both tech giants — plus newcomer London-based OneWeb — are racing to launch satellite internet services to offer global broadband connectivity, especially targeting remote and rural users.
The prize is a share of an expected $1 trillion market in the coming decades providing internet access to under-served communities worldwide. Morgan Stanley analysts see a potential for 300-500 million satellite broadband subscribers by 2040.
SpaceX as usual has moved fastest, growing Starlink’s customer base to a half-million subscribers in just two years since initial service launched.
Starlink’s network now spans 3,000 low-Earth orbiting satellites — giving substantially global coverage already with speeds up to 200 Mbps. Its Gen2 satellites aim boost speeds to 10 Gbps per user.
Amazon aims to leapfrog with an ambitious 34,000 km network of satellites in very low orbits from 350-590 km high. Kuiper says that will achieve sub-100 ms latencies needed for mobile applications.
Still, Project Kuiper is far behind, with no launches yet. Its initial goal was to launch the first two prototype satellites by end of 2022, which did not happen. It now aims to launch the initial five by 2024.
Race to Satellite Networks
| Company | # Satellites launched | Coverage | Speed | Latency | Users | Timeline |
|---|---|---|---|---|---|---|
| Starlink | >3,000 launched | Global | 200 Mbps (10 Gbps goal) | 20-40 ms | >500,000 | Providing service since 2020 |
| Project Kuiper | 0 launched | Goal of 98% global coverage | Up to 400 Mbps | <100 ms goal | 0 | Aims for 500 satellites up by 2026 |
| OneWeb | >70% of network up | Alaska to Europe service launched | 50 Mbps+ (200 mbps goal) | <50 ms | Unknown | Full global service delayed to 2023 |
The multi-launch SpaceX deal aims to jumpstart Kuiper’s deployment in second half of 2023 toward an initial goal of getting 578 satellites in orbit by 2026.
Longer-term, Amazon envisions a $10 billion network operating at 98% capacity globally. But to achieve that, Kuiper’s deployment really needs to shift into high gear and demonstrate rapid scaling, analysts say.
"Amazon needs to show meaningful commercial progress in 2023,” said industry analyst Tim Farrar of TMF Associates.
Challenges Ahead
Yet Amazon still faces considerable headwinds to scale up fast enough to become competitive and financially viable.
The huge capital costs ahead and long development timeline remains daunting. Morgan Stanley estimates Kuiper has likely spent $5 billion already just on R&D, design and early procurement, transmitters, and ground stations.
Going forward costs per satellite will drop as volume manufacturing kicks in. But the full 34,000 satellite constellation may still require upwards of $20 billion in capex, figures expense sharing firm ABC. That equals about 7% of Amazon’s total current assets — a sizable commitment.
Beyond getting rockets to launch them, expert analyst Quilty points to other major hurdles:
- Executing complex satellite production rollout
- Establishing worldwide ground station connectivity
- Securing spectrum access licenses globally
- Signing up wide industry adoption needed to support $500 terminals
- Competing against Starlink’s first-mover lead
On the plus side, Kuiper has hired a deeply experienced leadership team from across the aerospace industry. And it has technical strengths, like inter-satellite laser links to reduce long-distance signal lags.
Still, industry observers say profitability likely remains many years away. Competitors like Starlink and OneWeb similarly do not expect cash flow breakeven until later this decade at best.
“I would expect Kuiper will be lucky to hit breakeven in 10 years after it goes into service,” said analyst Caleb Henry. “This is an ultra-marathon, not a sprint.”
For now, much comes down to getting enough satellites up to establish initial coverage and proof-of-performance to begin commercial service as soon as 2024.
On that count, the SpaceX launches mark a key first step. Though still likely leaving Amazon hungering for the day Bezos’ own New Glenn rocket can shoulder a bigger share of the massive mega-constellation lift requirements ahead.
What’s Next?
Industry experts see Kuiper’s SpaceX launch deal as a stopgap while Amazon presses Blue Origin to accelerate efforts to get New Glenn certified and launching toward its 50-ton payload lift capacity. Quilty expects test launches could come sometime in 2024 or 2025.
Still, New Glenn’s heavy lift capacity would exceed Amazon’s nearer-term batch deployment needs. So Kuiper will likely procure smaller and medium-sized launchers for ongoing launches over the next five years.
Analysts point to startup launchers like Relativity Space’s 3D printed Terran R rocket that Kuiper has already reserved capacity on. Its 30-ton payload to orbit would allow launching 10 Kuiper satellites per flight — ideal for keeping assembly lines rolling.
But Terran R’s development timeline is also uncertain, not expected to launch until 2025. So Kuiper will need to secure slots on other emerging mid-sized launchers like SpaceX’s Falcon 9 rideshare missions.
The next year promises intriguing jockeying between Musk and Bezos’ rival space ventures. Yet both companies have huge ambitions that will require learning to partner amidst competition.
Friday’s SpaceX-Amazon launch deal “proves that these billionaire-backed companies can work together,” said industry analyst Chad Anderson of Space Angels. “Real progress will happen faster through cooperation.”
The rival internet satellite race has only just lifted off. But no matter how bitter the billionaire egos, the corporate pragmatism to start actually launching satellites matters most.
On that, Jobs’ Apple and Gates’ Microsoft likely offer the best analogy. Savage foes driving tech innovation faster through a mix of competition and coopetition.
To err is human, but AI does it too. Whilst factual data is used in the production of these articles, the content is written entirely by AI. Double check any facts you intend to rely on with another source.
