June 9, 2026

Dystopic Newsletter

Orbital Data Centers - the Technology Behind the SpaceX IPO Value Proposition

Orbital Data Center Based on High-Density Modular Compute Tiles (Thales)

​The SpaceX IPO is a whale, a monster, and the largest IPO on record. According to Investor’s Business Daily, in a June 3 SEC filing, SpaceX (SPCX) is offering 555.555 million shares at $135 per share in its initial public offering, raising $74.4 billion. The IPO is set to take place on Thursday, June 11, with the company debuting on the Nasdaq as SPCX the following day, Friday, June 12.

SpaceX is seeking an initial IPO valuation of $1.77 trillion, making it the largest IPO in history (see graphic). However, there is controversy in the financial world over that valuation, with several financial analysts, including Morningstar, believing SpaceX should be valued below $1 trillion.

As I reported in the May 27, 2026, Dystopic News Letter, SpaceX Incredible Company - Risky IPO & World Flash Point Update: The SpaceX IPO and business plan are only partially focused on SpaceX Space Operations; the real focus is on a risky bet on AI and space-based data centers.

What technologies and risks are associated with Orbital (Space-Based) Data Centers? Are there competitive alternatives? That is the focus of this week's Dystopic newsletter, based in part on my technology briefing: Orbital Data Centers – Logistics, which I delivered on May 27 at the SmallSat Europe 2026, held at the RAI conference center in Amsterdam.

Enjoy video content? You can view the lecture HERE​

Orbital Data Centers – The Technology, Risks, and Logistics of Massive Computing Power in Space

Deploying data centers in orbit offers radical benefits. All the major space transport and AI infrastructure players, including SpaceX, Amazon/Blue Origin, and Google, are planning constellations of 1000s, even 100,000s, of satellite data centers in orbit. The logistical hurdles of getting them there and keeping them running are immense. To understand this, we must examine the physical infrastructure needed for space-based servers, including thermal management, power generation, and payload integration.

As we have previously noted, SpaceX's IPO is based on the feasibility and inevitability of orbital data centers being economically superior to terrestrial data centers.

Let's start with the perceived benefits; what are the motivations and advantages of placing data centers in space?

Why Space? Orbital Data Centers vs Terrestrial Data Centers

The headlines are full of stories about towns and localities protesting or banning the construction of new AI terrestrial data centers. The reasons are many, but particular concerns include the use of water resources, the cost of building new power plants, and the inevitable rise in electricity rates for consumers located near the data centers.

Given these issues, it’s no wonder that in early 2026, SpaceX’s request with the FCC to build out orbital AI data centers drew so much attention. The grand scope of the request for a satellite constellation of up to 1,000,000 satellites caught the financial community and the world’s attention.

As the table comparing orbital and terrestrial data centers illustrates, while multiple factors are at play, energy availability and scalability are the primary factors favoring orbital data centers.

• Energy Availability: Continuous, uninterrupted solar power based on placing satellites in polar “dawn-dusk” orbits (polar orbit), providing gigawatts of power at a $0.01 per kWh. Five to fifteen times less expensive than terrestrial power plants at $0.05 to $0.15 per kWh
• Scalability: Space is vast, with no land constraints. However, at some point the end-of-life issue for data center satellites comes into play. The obvious terrestrial counterpoint is that locations for terrestrial data centers are extremely limited unless we consider deployment on the ocean floor.

Orbital data centers face two key issues. First is the cost of deploying satellites. If SpaceX can’t achieve a launch cost below $200 per kilogram, orbital AI satellites are not financially feasible. That means SpaceX's bet on Starship must pay off. Second, the AI chips used today are not designed for space and are not built to conserve energy. They are designed for maximum compute power and, in today's implementations, make no provisions for energy conservation. Error recovery is handled in software by commercial AI ASICs rather than in dedicated hardware, thereby reducing the overall AI system's compute capacity.

Using advanced ASIC design techniques to reduce power consumption while maintaining performance gives SpaceX and other orbital datacenter constellation companies a significant edge over their terrestrial competitors.

Who are the key stakeholders with plans to build orbital data centers?

As it turns out, SpaceX is not alone. There are a number of companies vying to compete with SpaceX's vision of placing AI computing power in orbit. Amazon, with its subsidiary rocket company Blue Origin, is developing Project Sunrise, a constellation of over 51,000 satellites. Google, in conjunction with Planet Labs, is conducting a series of experiments called Project Suncatcher to test the viability of its TPU AI ASIC. VC-backed company, StarCloud, has filed plans with the FCC for an 88,000-satellite constellation.

Of course, numerous startups, such as Sophia, Space Axiom, Space, and Aetherflux, are being funded to develop component and subsystem technologies to build out these planned constellations. More are sure to come.

These are massive satellite constellations. The plans outlined in the FCC filings are two orders of magnitude larger than the current mega-constellations being deployed by SpaceX and Blue Origin. Consider the logistics of SpaceX's million-satellite constellation compared with Starlink, which has just over 10,000 satellites in orbit. These are massive undertakings, and the logistics and operations to achieve them will be difficult, but not impossible.

Consider this fact. In 2025, SpaceX dedicated 99 of its 165 launches to deploying Starlink. The Starlink constellation contains roughly 11,000 satellites. Rounding the numbers a bit, 100 launches using the Falcon 9 SpaceX launch vehicle are required to support 10,000 satellites. If SpaceX were limited to Falcon 9 launch vehicles, 10,000 launches would be required each year to maintain the constellation. That is roughly 28 launches per day, with a cost to LEO of roughly $ 2,000 per kilogram.

Enter SpaceX Starship, capable of lifting 100 metric tons to orbit, six times the launch capacity of Falcon 9 at a tenth thet cost at $200 per kilogram to LEO. That translates to fewer than 1,700 launches per year, or 4.5 launches per day. That would still be an incredible logistical feat.

Starship itself is not the only logistical improvement, as we will discuss later in this Newsletter, in-orbit servicing make addition reductions in the required number of launches once the constellation is in place

The Obsolescence Factor: AI Data Center Compute Life Cycle

If we are to put constellations of data center satellites into orbit, it's important to understand the lifespan of an AI computing blade, the smallest replaceable part in the computing racks that make up a data center.

The following diagram illustrates the typical lifespan of a data center blade, approximately 5 years. The blade is deployed and then operates for 1 to 3 years in prime mode, handling the heaviest compute loads, typically LLM training, before it is repurposed and replaced by a new generation of computer blades. For another couple of years, the blade has a secondary life performing less demanding compute tasks until it is decommissioned.

Every two to three years, Moore's Law, which describes the semiconductor industry's ability to double the number of transistors while cutting the cost per transistor in half, renders a two- to three-year-old computer blade practically obsolete. In some cases, such as the transition from Nvidia’s Blackwell chipset to the Vera Rubin chipset, the performance improvement is so compelling that the blades were replaced well before their typical five-year lifespan. This faster replacement has been termed the "Nvidia effect."

For orbital data centers to remain competitive, they will either have to be completely scrapped and replaced or be capable of in-orbit upgrades to maintain their competitive edge over terrestrial data centers. For SpaceX's million-satellite constellation, 200,000 satellites will need to be upgraded each year for as long as the constellation is intended to operate if in-orbit updates are not possible. How does that make financial sense?

This raises a question. When we decommission each satellite, will it be de-orbited to burn up in Earth's atmosphere or boosted into a junkyard parking orbit? We worry about water and electrical generation pollution from terrestrial data centers. Orbital data centers have an equivalent pollution factor in the form of space junk. A solution will be required within a few years of deploying these massive orbital data center constellations.

We don’t have an answer for the disposal problem yet. Let’s move on move on explore Orbital Data center design

Orbital Data Center Design

Three factors drive Space Data Center Design:

• AI ASIC power consumption: AI ASIC chipsets over the next year will move from 1.5 kW per chipset to nearly 3kw per ASIC. Essentially doubling
• Solar Panel Area: State-of-the-art space-based solar panels generate between 400 and 600 watts per square meter. In addition, solar panels act as a sun shield for the radiator, which are criytical to cooling the AI ASICs so they can operate.
• Radiator: Cooling in space is based on infrared radiation. Behind the shade provided by the solar cells, space is a cool -270 degree C, perfect for radiative cooling. Assuming the radiators are conducting heat for the AI Asics at 27 deg C, a single-sided radiator is capable of dissipating 450 watts per square meter

As it turns out, the area of solar panels used to generate power is slightly larger than that of a single-sided radiator needed to maintain cooling for that power consumption. This is ideal because the areas needed are roughly equal. There will be no excess waste of material due to an imbalance between power generation and power dissipation. Each kilowatt of power required by an AIASIC drives 2.5 square meters of solar area and 2.25 square meters of a single-sided radiator.

In practical terms, a single Nvidia VERA-Rubin chipset module consuming 2.3 kilowatts requires nearly 6 square meters of solar cell/radiator area.

The size and mass of an orbital data center satellite are dominated by Solar Cell / Radiator Area

So, what will the Design and architecture of space-based data centers look like? There are essentially two schools of thought:

• Distributed "Compute Constellations" ( Clustered - Mesh Networks)
• High-Density Modular "Compute Tiles” (Mega Space Structures)

The Distributed "Compute Constellations" architecture is based on the evolution of the payload fold-out antenna architectures we see today in SpaceX and Amazon Leo communications satellites. Groups of satellites would fly in orbit together as clusters connected by laser crosslinks. Multiple clusters per orbital plane and multiple orbital planes, with a mesh network of laser data links between clusters and from each cluster to the earth-based ground stations, create the entire constellation.

This architecture is favored by SpaceX, Amazon/Blue Origin, Google’s Suncatcher, and VC-backed Starcloud (Stem & Leaf). In fact, SpaceX’s FCC proposal included over 26 sun-synchronous orbits, with 50-kilometer altitude spacing, ranging from 700 km to 2000 km.

The second architecture is based on the concept of a High-Density Modular "Compute Tile." Each tile is designed as an integrated unit comprising an AI ASIC(s), a solar panel, and a radiator. The fairing of the designated launch vehicle (with fairings between 5 and 9 meters in diameter) would set the maximum tile size between 20 and 63 square meters, depending on the launch vehicle used.

Tiles would be directly connected via optical fiber into a mega structure containing 10s to 100s of tiles. Thrusters to maintain orbit and optical ground communications would be distributed across each mega structure. Multiple nodes on multiple orbiting planes make the constellation.

This architecture is proposed by two start-up companies, Sophia Space and Axiom Space. These companies can begin by deploying single tiles in orbit, either standalone as a data center function or combined with radar or optical Earth observation satellites, to perform edge AI processing. This is likely where the tile architecture will excel, because neither Sophia Space nor Axiom Space has a solution for building their proposed megastructure in space.

It is likely that both architectures will be deployed. Tiles for edge AI computing and Distributed "Compute Constellations" will serve as direct replacements for many workloads performed by terrestrial data centers. In either case, all satellites will experience connectivity latency compared with terrestrial data centers and will need to share and distribute workloads among edge AI processing, orbital data centers, and Earth-based data centers. In short, the entire system will be a Hybrid Earth-Space Architecture.

In the Hybrid Earth-Space Architecture, Edge AI will operate in standard LEO orbits, with most of the workload focused on the pre-processing of Earth observation data. Non-latency-sensitive workloads would be performed in SPACE, while latency-sensitive tasks would be assigned to EARTH-based data centers.

With the exception of edge AI application data centers, satellites will be placed in sun-synchronous polar orbits, so-called “dusk-dawn” orbits, where the satellites are always illuminated by sunlight. A majority of data center constellations will be orbiting above 700 km. Unlike LEO satellite deployments at 500 km or less, deployments at 700 km or greater offer several benefits, including longer orbital life and reduced fuel consumption for station keeping.

Typical LEO satellites fall out of orbit in 5 years – roughly equivalent to the lifespan of an AI chip set (Is that a good thing – probably not). At higher orbits, satellite life cycles are greatly extended. At 700 km, a satellite can typically orbit for 25 years. At 2000 km, the orbital life cycle extends to 1000s of years.

Keep this in mind as we continue our discussion. Let's take a closer look at our power-hungry AI ASIC chipsets

AI ASIC Power Consumption: The Driver of All Orbital Data Center Design

The power consumption of AI ASIC chipsets drives all design decisions for orbital data-center satellites. Consider the power consumption trends for each generation of Nvidia’s flagship AI chipsets. Power consumption for each new generation of chipsets is nearly doubling, at approximately 40% per year. At 2.3 kilowatts per “blade,” Nvidia’s latest-generation chipset, Vera Rubin, is hitting a “Thermal Wall.” Air cooling is no longer feasible, and direct fluid cooling techniques are required. Temperatures high enough to damage the ASICs can occur during normal operation. To avoid damage, thermal throttling, which reduces the clock rates and processing power of the ASICs, is used to keep the system below its maximum operating temperature.

Power is increasing despite improvements in silicon process nodes under Moore's Law. NVIDIA and AMD have focused their engineering on performance at the expense of power. AI ASICs are expanding horizontally, both to increase processing power and to assist in thermal dissipation. This trend is leading to massive wafer-scale processors. Take Cerebras Systems as an example.

Their Wafer-Scale Engine (WSE-3) measures 46,225 square millimeters (8.5 x 8.5 inches), roughly the size of a standard dinner plate. A single WSE-3 chip nearly spans the entire area of a 12-inch (300 mm) silicon wafer (hence the term “Wafer-Scale”). To put its size into context, it contains 4 trillion transistors, 900,000 AI-optimized cores, and is 57 times larger than an Nvidia H100 GPU.

Of course, NVIDIA and AMD are not the only solutions. Intel, SpaceX/Tesla, Google, and Amazon have significant multi-generation AI ASIC programs. Each of these programs is based on custom features tailored to their specific AI workloads and deployments. We would expect that in the near future, SpaceX/Tesla and Amazon will be motivated to use advanced ASIC architectures not only for error detection, error correction, and self-healing techniques but also to save significant amounts of power. Of particular interest R2 semiconductors, originally designed to optimize power consumption in smartphone radio processors and RF power amplifiers, that are now being applied to AI ASIC design.

The bottom line is that AI ASICs are hitting a thermal wall, and the current mainstream processors are diverging from what is needed for space-based data centers. Power-saving techniques will be critical to the success of orbital data centers.

Remember, each kilowatt of power drives 2.5 square meters of solar panel area and 2.25 square meters of single-sided radiator area. Imagine the savings in weight and launch cost if we could halve AI ASIC energy consumption?

Let’s turn our attention to radiators – we have a lot of heat to dissipate.

Space-Based Radiators

Cooling in space relies on infrared radiation. As we previously discussed, satellite data center design uses solar cells to both generate power and serve as a sun block, providing shade for the radiators (-270 deg C).

The International Space Station has been a proving ground for thermal control for the better part of two decades. NASA prototyped numerous systems to evaluate effectiveness over the decades. State-of-the-art space-based radiator uses heat pipes filled with either Aluminum-ammonia or copper-water working solutions to distribute heat across the radiator structure. Modern/Lightweight (Carbon-Carbon) radiators at 2 kilograms per square meter weigh a third of their metal counterparts and can dissipate 450 watts per square meter.

There are two basic designs for a heat pump system: passive or active

Passive Heat Pipes are based on a capillary wick structure to operate in zero gravity

• Advantage: Reliable, long life, no active parts, minimum weight, no power load
• Disadvantage: Designed for a fixed thermal load (not variable). As it turns out, given a fixed array of solar cells, their thermal load (and that of the supported AI processors) will also be fixed.
• Oscillating Heat Pipes (OHPs) are especially promising passive radiators that improve weight and performance over earlier designs

Active Heat Pipes are based on pumps to move the working fluid

• Advantage: Allows for variable load, more efficient, higher thermal load
• Disadvantage: complex with moving parts, lower reliability, energy use
• Loop Heat Pipes (LHPs) designs are Capable of transferring higher heat loads over longer distances with high efficiency

Ideally, orbital satellites would use passive heat pipes. If active heat were required, a redundant backup pump would be necessary, as would a simple plug-and-play in-orbit replacement capability.

So … let's take a look at in-orbit servicing:

Orbital Life Cycle: A case for modular design in-orbit servicing

There is a very strong case for modular orbital satellites. Consider the following facts:

• A majority of orbital data center satellites will be in polar orbits between 700 and 2000 Km. At these altitudes, a satellite can remain in orbit for at least 25 years before orbit decay and far longer if the satellite can be refueled in orbit
• Solar Cells and Radiators have a lifespan of up to 25 years in orbit, far longer than a typical AI data center chipset/blades
• AI CPU/Blade, Free Space optics, and station-keeping propellant have shorter lifespans, typically 5 years.
• At least 80% of the mass of any orbital data center is the solar cells and radiators.

Clearly, a modular orbital data design using solar cells/radiators as the backbone of the system could reduce decommissioning waste and launch costs if it enabled a 5-year replacement cycle and an in-orbit replaceable server farm module could be upgraded five times over its service life.

As the old datacenter module it is replaced, it would be decommissioned either by deorbiting to burn up in Earth’s atmosphere or by burning the remaining propellant to place the module in a parking orbit. The replacement module would upgrade the CPU, free space optics, recharge propellant, and radiator working fluid.

A modular replaceable architecture could conceivably reduce the mass to orbit life cycle costs by roughly 64%

While our focus has been on the data centers themselves, consider Astroscale, a pioneering Japanese space startup currently valued at approximately $1.7 billion on the Tokyo Stock Exchange and trading over-the-counter (OTC) in the US under the symbol ASRHF. Astroscale is the world's first publicly traded company dedicated entirely to orbital sustainability, active debris removal, and in-space satellite servicing. If space-based data centers take off, companies like Astroscale will be well-positioned for significant growth.

Speaking of growth, the stratospheric valuation of the SpaceX IPO is due, in part, to its degree of vertical integration. What does that mean? Read on …

The Vertical Integration Advantage

Early in this newsletter, we established that launch costs below $200/kg are required for economically viable space data center deployment. In the near term, SpaceX Starship and the next-generation Blue Origin (despite last week's setbacks) are the only launch vehicles that can reduce future launch costs by 8x from current levels and reach the magic 200/kg threshold.

Both SpaceX and Amazon/Blue Origin can leverage their “full-stack” capabilities to deliver the space-based data center vision:

• Launch vehicles and launch cadence build out their respective constellations – scaling from a Falcon 9 launch every 2 weeks to four to five Starship launches a day
• Exiting satellite communications and ground station network Starlink & Amazon LEO, including: Terrestrial Data Centers, established AI businesses/partnerships, and Mature, well-funded custom AI chipset development programs

Despite these leveraged capabilities, significant opportunities exist across the ecosystem and in niches for other companies:

• Chipsets and Foundries: TSMC, Intel, Nvidia, AMD, etc.
• Earth Observation edge compute AI (Nvidia Jetson-Orin & Thor, & alternates)
• Full-stack supply chain components – optical terminals, thrusters, etc
• In-orbit servicing - Astroscale
• Alternate launch services: Rocket Labs, Firefly, etc.

Given the difficulty of launching vehicles (yes – rocket science is very challenging! Consider last week's Blue Origin debacle), is it any surprise that Google/Alphabet is rumored to be in talks with SpaceX for launch and communications services for its own space-based data centers?

Based on the progress and success, while a business plan for orbital data centers is technically risky, SpaceX is at the forefront of executing the vision laid out in the S-1 document to go public.

SpaceX has one last capability in its favor – its own foundry and ASIC designs. Let's call this "The Terafab factor."

Custom Space-Based AI/CPU – The Tarafab Factor

Today, AI satellite prototypes are testing existing commercial core and edge AI ASICs. They are designed for performance, not power savings. They attempt to mitigate space radiation effects through physical shielding and software error detection and correction, not through ASIC and memory design.

Ideally, the memory and AI ASICs for orbital data centers differ from their conventional counterparts in two distinct architectural design features: hardened, self-healing system design and advanced clocking and power control.

Space is not a forgiving environment for ASICS, Memory, and electronics in general. cosmic radiation, solar flares, Coronal Mass Ejections (CMEs), and solar radiation storms (energetic), despite the protection of the Earth's magnetic field, satellites in LEO can experience a number of issues.

• Single Event Effects (SEE): a single high-energy particle can create a non-destructive bit error in the system memory or ASIC(s)
• Multi-Bit Upset (MBU): a single strike with multiple failures and ECC (Error Correcting Code) failure
• Single Event Latch-up (SEL): an energetic particle creates a destructive hardware failure
• Total Ionizing Dose (TID): cumulative damage from long-term exposure – causes power leakage

Custom AI ASICs with advanced fabrication, silicon and packaging architectures, and system-wide ECC, self-test, and self-healing offer significant improvements in power efficiency and resilience to radiation effects compared with conventional designs. Consider the R2 silicon imposer and active distributed power regulation solutions licensed by both Intel and Samsung. The R2 solutions offer roughly 50% lower power consumption than standard design practices.

Consider the following: SpaceX is investing nearly $130M in Terrafab, its massive silicon foundry in Texas. Terrafab has teamed with Intel to roll out Intel‘s advanced A14 (1.4nm) process technology. The Intel A14 process, combined with the R2 design solution, would double compute power for a given AI satellite power configuration while also enabling hardware error detection, error correction, self-test, and self-healing for robust orbital operation.

There is a method to the madness of the Terafab investment. SpaceX (along with xAI) will create an incredible technical advantage against its competitors .

Some final thoughts on the SpaceX IPO and space-based data centers

At first glance, a business plan based on orbital data centers appears incredibly risky. However, it is technically feasible. Heavy launch vehicles hiting $200/kg to LEO orbit, modular satellite design, in-orbit servicing, custom silicon, and a fail-fast culture willing to spend the money have constant innovation through iteration are all elements that greatly reduce valuation risk of the SpaceX IPO.

As it turns out, a $1.8 trillion valuation is not so crazy considering the obstacles facing the build-out of terrestrial data centers.

There is a terrestrial solution that could pose a threat to SpaceX’s grand plans. A viable alternative could be floating or underwater data centers based on:

• Seawater Cooling: Instead of energy-hungry chillers, subsea pods circulate cold ocean water through heat-exchange radiators. The ambient sea passively absorbs heat, bringing Power Usage Effectiveness (PUE) down below 1.15 in some deployments
• Modular Reactors (SMRs): To bypass congested, fossil-fuel-dependent land grids, engineers propose pairing these centers with offshore Small Modular Reactors (SMRs) or floating nuclear plants. This provides continuous, carbon-free baseload power to the thirsty Artificial Intelligence (AI) and cloud computing hub

The Earth is 75% coverd by ocean, so there would be plenty of room to scale.

https://www.scientificamerican.com/article/china-powers-ai-boom-with-undersea-data-centers/

As Scientific American reports, China is already experimenting with submerged seawater-cooled data centers off the coasts of Shanghai and Hainan, powered by ocean currents and nearby wind farms rather than SMRs.

There are a number of technical and environmental issues to be addressed:

· Maintenance & Upgrades: Data centers will operate in “inner space,” a world well understood by nuclear submariners. The hardware is sealed underwater, so fixing broken components or upgrading server blades requires either a staffed resupply/maintenance submersible (yes, those exist in the U.S. Navy today and could be adapted) or retrieving the entire module, much like we have discussed for orbital satellites.

· Ecosystem Impact: Researchers are evaluating the long-term effects of discharging heated water into the ocean.

So the economic threat to deployment in outer space could be deployment in inner space (underwater). Both have technical hurdles, yet both are feasible.

From Triumph to Tragedy – the Blue Origin Launch Catastrophe and Repercussions for NASA’s Lunar Ambitions

A tale of triumph turns to tragedy for Blue Origin. As I reported in the November 24^th, 2025 Dystopic Newsletter:

Just over 6 months later, that triumph turned to tragedy for the integrated space launch company and SpaceX competitor. As reported by the BBC, Blue Origin's fourth mission, NG-4, which was set to launch 48 satellites for Amazon's Leo broadband network in early June, exploded on launch pad LC-36 at Cape Canaveral. LC-36 is the only facility in the world built to launch the New Glenn rocket. Until the launch pad is rebuilt and re-certified, Blue Origin has no way to fly its largest rocket, and analysts expect that to take months, not weeks.

It could take longer than that. In 2016, a SpaceX Falcon 9 exploded at Space Launch Complex 40 during preparations for a static-fire test. SpaceX took 15 months to rebuild the pad. Fortunately, the damage was limited. During a live news conference, Blue Origin CEO Dave Limp indicated that the explosion’s damage was not as extensive as it first appeared.

Later, in a post on X, Limps stated optimistically, “We will fly again before the end of this year.”

We can only hope the Blue Origin CEO is correct. According to Space News, the explosion has ramifications across NASA and commercial space projects:

• NASA had just awarded Blue Origin contracts to deliver lunar rovers being developed by Astrolab and Lunar Outpost
• Amazon LEO planned to use New Glenn to deploy more than a third of the total Amazon LEO constellation of 3,232 satellites. Amazon LEO has faced several program delays, and this accident will further delay the constellation's deployment.
• AST SpaceMobile, whose direct or device technology is used by AT&T, Verizon, and a number of international cellular carriers, had plans to launch 45 satellites by the end of the year using New Glenn communications. It is now scrambling to find launch alternatives.

Before we wrap up this week’s Dystopic lets take a deep dive into LC-36 and the New Glenn Launch vehicle.

New Glenn Launch Complex (LC-46)

Launch Complex 36 (LC-36) is located at Cape Canaveral Space Force Station, just nine miles (14 km) away from Blue Origin’s rocket factory. Blue Origin invested more than $1 billion to rebuild the launch site from the ground up. Completed in 2021, LC-36 is the first newly rebuilt launch complex since the 1960s.

In addition to the launch pad, the complex is home to Blue Horizon’s vehicle integration, first stage refurbishment, propellant facilities, and an environmental control center. LC-36 is the former home of more than 140 Atlas II/III launches, including the Mariner, Pioneer, and Surveyor missions.

New Glenn Heavy-Lift Vehicle

According to Blue Origin’s published specifications, New Glenn is a heavy-lift, partially reusable launch vehicle developed by the Company. Like the SpaceX Falcon 9, New Glenn is designed to carry high-volume commercial and government payloads. The standard two-stage rocket stands 98 meters tall and features a 7-meter payload fairing. This falls between the 5.2-meter Falcon 9 and 9-meter Starship fairings used by SpaceX.

The first stage is equipped with six hydraulically actuated landing legs to enable vertical landing and reuse. It is designed to be recovered at least 25 times and to land on the massive sea-based platform named Jacklyn (see the image at the start of this section).

New Glenn is capable of launching 45,000 kg into Low Earth Orbit (LEO). In comparison, SpaceX Starship is designed to carry 100 kg to LEO in reusable mode, Falcon 9 17,500 kg, and Falcon Heavy 57,000 kg.

For you space fanatics, here are a few additional specs for what is designated as New Glenn 7 x 2 configuration:

First Stage

• Engines: Seven BE-4 liquid oxygen (LOX) and liquefied natural gas (LNG) engines
• Thrust: Over 3,800,000 lb-ft (17,100 kN) at liftoff
• Landing Gear: Features four movable aerodynamic control fins and strakes for flight stabilization and vertical landing.

Second Stage

• Engines: Two BE-3U liquid hydrogen (LH2) and liquid oxygen (LOX) engines
• Thrust: Over \(320,000\) lb-ft of vacuum thrust

Blue Origin is developing a new super-heavy-class launch vehicle, designated New Glenn 9x4, to better compete with SpaceX Starship. As the name suggests, New Glenn 9x4 has nine BE-4 engines (liquid oxygen and liquefied natural gas) in the first stage and four BE-3U engines (liquid oxygen and liquid hydrogen).

It is unclear how the LC-36 explosion will affect New Glenn 9x4 development. One thing is clear: this failure will drive more business to SpaceX in the interim and certainly bolster the SpaceX IPO. For a while, SpaceX will have the commercial space launch market cornered. (That’s an exaggeration - Rocket Labs is an option for medium and smaller payloads.)

That’s a wrap for this week …

Dystopic- The Technology Behind Today's News

Thank you for your readership and support. Please recommend Dystopic to friends and family who are interested, or just share this email.

Not on the Dystopic mail list? New Readers can sign up for Dystopic HERE ​

If you have missed a Dystopic News letter, you can find select back copies HERE ​

Dystopic is brought you by Paul Struhsaker, the author of: HOW THE HELL DID WE GET HERE? A Citizen’s Guide to the New Cold War and the Rebuilding of Deterrence

A #1 new release in National and International Security books on Amazon!

Available on Amazon: https://a.co/d/1FKGJtB or order from your favorite book seller.

See you next week!

Follow Me on Social

​​​

​Unsubscribe | Update your profile | 113 Cherry St #92768, Seattle, WA 98104-2205