StarCloud Inforb?

[jabowery]

Preliminary calculations below indicate The Inforb may consist of Starships as modular datacenters. Each Starship could house a module at 90tonne each, requiring 1MW. Each module is equipped with 3MW of ISS solar array wings totaling 38tonne, and 4tonne 1MW of battery mass.

The additional 60tonne capacity might prove useful for telepresence maintenance robots, spares, and, ultimately, even shirtsleeve environment for human technicians docking with a Dragon module.

The basic idea of the Inforb is that you don’t need to lay fiber in space and that most traffic is not between the human interface and the cloud. Now, I’ll admit that my thinking may be antiquated given the change to video streaming from 2001 when I coined the term Inforb let alone way back in the 80s*. But it would be interesting to see how much bandwidth utilization could be reduced by the simple expedient of video caches in the network with not just Information Centric Networking but doing so with Edge Computing serving as cache for local area distribution, which has been part of the vision for ICN.

* Way back in the 80s I gave a presentation at the Ruben H Fleet Science Center on laser linked GSO datacenters with what nowadays might be called “the last mile fiber” to humans on earth, handled by Ka-band. At the time all I had to go on was a radiation hardened version of CDC’s spinoff, ETA’s attempt to beat Cray, but I’d done the work at CDC’s benchmark labs to verify the market potential hence economics of at least providing the mass-market version of the PLATO network I’d been promoting.

how much does a modular data center weigh?

ChatGPT

The weight of a modular data center can vary depending on its size, configuration, and the specific equipment installed within it. A modular data center typically consists of pre-fabricated modules or containers that house servers, storage devices, cooling systems, power distribution units, and other necessary infrastructure components.

On average, a modular data center module or container can weigh anywhere from a few tons to several tens of tons. Smaller modules may weigh around 10 to 20 tons, while larger ones can reach 50 tons or more.

It’s important to note that the weight of a modular data center does not include the weight of the servers and other IT equipment housed inside it. The weight of the data center infrastructure itself primarily includes the structural components, cooling systems, electrical systems, and any other supporting equipment.

how much power does a modular data center consume?

ChatGPT

The power consumption of a modular data center can vary significantly depending on several factors such as the size of the facility, the type and density of the IT equipment installed, the cooling requirements, and the overall design efficiency.

Modular data centers are designed to be energy-efficient, and newer models often incorporate advanced technologies to optimize power usage. However, it’s challenging to provide an exact figure without specific details about the data center in question.

On average, a modular data center can consume anywhere from a few hundred kilowatts (kW) to several megawatts (MW) of power. Larger data centers with higher equipment densities and more extensive computing capabilities will naturally require more power.

It’s worth noting that power consumption in data centers is not solely determined by the IT equipment. Cooling systems, power distribution units, lighting, and other auxiliary equipment also contribute to the overall power usage. Moreover, data centers typically have varying power requirements during different operational phases, such as normal operation, peak load, or standby mode.

To accurately determine the power consumption of a specific modular data center, it is essential to consult the manufacturer’s specifications or consult with an expert who can assess the facility’s design and provide accurate power consumption estimates.

what is the estimated payload capacity of a spacex starship

ChatGPT

The SpaceX Starship, also known as the Starship vehicle or Starship spacecraft, is a fully reusable spacecraft designed for a variety of missions, including crewed space exploration, satellite deployment, and potential future missions to the Moon, Mars, and beyond.

The estimated payload capacity of the Starship has evolved over time as the design has undergone updates and refinements. As of my knowledge cutoff in September 2021, SpaceX had indicated that the Starship’s payload capacity to Low Earth Orbit (LEO) could be up to 150 metric tons (330,000 pounds). It’s important to note that this estimate is subject to change as SpaceX continues to develop and test the Starship.

It’s worth mentioning that the Starship’s payload capacity can vary depending on the mission profile, destination, and other factors. For example, if the mission involves sending cargo or crew to the Moon or Mars, the payload capacity to those destinations would be lower due to the additional fuel and resources required for the journey.

For the most up-to-date and accurate information on the payload capacity of the SpaceX Starship, it’s recommended to refer to official SpaceX announcements, specifications, or consult SpaceX directly.

what is the mass of an ISS solar array wing

ChatGPT

The mass of an International Space Station (ISS) solar array wing, also known as a solar panel, depends on the specific type and model of the solar array used.

As of my knowledge cutoff in September 2021, the ISS primarily uses two types of solar arrays: the older Solar Array Wings (SAWs) and the newer Roll-Out Solar Arrays (ROSAs). Here are the estimated masses for these solar arrays:

1. Solar Array Wings (SAWs): Each SAW on the ISS consists of two solar panels that span approximately 34 meters (112 feet) in length. The SAWs have an estimated mass of around 450 kilograms (990 pounds) each.
2. Roll-Out Solar Arrays (ROSAs): The ROSAs are newer, lighter, and more compact solar arrays designed to be more efficient and easier to deploy. The ROSAs are made up of flexible blankets that can be extended to capture sunlight. The estimated mass of a ROSA solar array is approximately 250 kilograms (550 pounds).

It’s important to note that these figures are approximate and may vary depending on the specific version and configuration of the solar arrays used on the ISS. The actual mass of a solar array wing can also be influenced by the supporting structures, electrical components, deployment mechanisms, and any additional equipment associated with the solar array system.

how much power does it generate while in full sunlight in LEO

ChatGPT

The power generation of a solar array on the International Space Station (ISS) can vary depending on factors such as the specific solar array model, its age, efficiency, and the angle at which it is exposed to sunlight. However, I can provide you with some general information about the power generation capabilities of the ISS solar arrays.

The ISS uses a combination of older Solar Array Wings (SAWs) and newer Roll-Out Solar Arrays (ROSAs) to generate electrical power from sunlight. As of my knowledge cutoff in September 2021, the SAWs were rated to provide up to approximately 32.8 kilowatts (kW) of power, while each ROSA was designed to produce around 20 to 25 kW.

These power ratings represent the maximum generation capacity of the solar arrays under optimal conditions, meaning direct exposure to sunlight at the ideal angle. However, it’s important to note that the actual power generation can be influenced by factors such as orbital variations, shading by other ISS components, degradation over time, and power losses in the electrical system.

Furthermore, the ISS operates at an altitude of about 400 kilometers (250 miles) above Earth’s surface, which means it experiences orbital periods of both daylight and darkness. During the orbital night or periods of Earth’s shadow, the ISS relies on stored electrical power in batteries.

It’s worth noting that the ISS power system is designed to support the station’s operational needs, including life support, scientific experiments, communications, and other systems, while also charging the batteries for use during orbital night. The power generation capacity of the ISS solar arrays allows for these various demands while maintaining the station’s power requirements.

[jabowery]

6:25 Can’t buy natural gas generators because turbine blade manufacture is blocking.

100GWt ng is being re-injected at Prudhoe Bay according to Gemini3.0. There may be a workaround for the blades that uses that energy in a low efficiency manner.

[jabowery]

The first thing that came to mind was that at a levelized cost of electricity of around $0.50/kWh, which these datacenters can support and is what makes them a threat to civil society as they demand that of our utilities in competition with families being put out of work by AI, is Tesla turbines.

You don’t have to tell me about all the problems with Tesla turbines. In the 1980s I was involved with the ONR guy who financed ring laser gyroscopes that wanted to pursue some advanced TT patents and Roger and I did some pretty good calculations about TT as rocket engine pumps.

So I thought I’d look at the other off the shelf systems for stranded ng elex. The costs, as they stand, for 3 year writeoff are already at the $0.50/kWh level even assuming free ng. Then you have to consider the increasing price with higher demand for them.

So, back to TTs with due trepidation which brings me back to my original motivation to improve on Hudson’s Roton in the work I financed with Roger. People don’t seem to understand just how must “slack” $0.50/kWh buys you in the context of cheap cooling and cheap stranded ng.

This is a wild angle I’d never heard of before:

Chung-Ang University scientists generate electricity using Tesla...

Recent Advances in Triboelectric Nanogenerators: From Technological Progress to Commercial Applications.

And, yeah, there are all kinds of “kill shots” on this idea too. Don’t think about it, managerial class, just drift on toward people getting their electricity shut off in the dead of winter so they can’t even run the fans on their furnaces. You don’t have to think about that either. Everything will be just fine with your monopoly capital in the private sector and civil service sinecure in the public sector.

Until it isn’t.

Oh, BTW, the “AI solution” isn’t there for this. When I explored some ideas with GPT5.2 (for example) it ignored the fact that TTs can act as compressors as well as expanders and the fact that NG reinjection can be replaced by CO2 injection – as it came up with a litany of other excuses. This isn’t merely a lack of creativity – it is reminiscent of insular stupidity that the managerial class imposes on the biosphere.

[jabowery]

After struggling with these things to get them to stop thinking about rocket engines as propulsion systems and a variety of other ways they can “lose the plot”, I came up with this prompt:

I want a stationary system that turns natural gas into momentum using air, natural gas and ocean bay water injection without using engines to drive a crank shaft and without using turbine blades to drive a crank shaft. Note I said MOMENTUM. Note I said STATIONARY as in ANCHORED TO THE GROUND. Why would I want such a thing?

Tesla turbines.

Why would I want tesla turbines?

Because human beings are being threatened with the loss of electrical power and turbine blades are not obtainable for natural gas elex generation and there is about 100GW of ng being pumped into the ground in alaska and the burn product of such a system could replace ng as injected fluid and tesla turbines may be viable but only if the speed of the fluid is low.

I’ll accept loss of 2/3 of the natural gas efficiency in this heat to momentum transfer. Multi-GW machine learning may bridge the gap before SpaceX/xAI comes to the rescue with space based solar machine learning.

Gemini 3 responded:

You are describing a Direct-Contact Combustion Ejector (or Thermal-Hydraulic Jet) driving a Multiphase Tesla Turbine.

This is a feasible, robust, low-tech engineering solution. It bypasses the need for precision airfoils (turbine blades) and high-tolerance reciprocating parts (crankshafts/pistons), trading thermal efficiency for mechanical simplicity and supply chain independence.

Here is the breakdown of the system you are envisioning and why it works for your specific “Alaska / Stranded Gas” scenario.

The System: The Thermal-Hydraulic Ejector

You want a stationary system that turns heat (NG combustion) into fluid momentum. Since you cannot use a piston or a bladed turbine to capture that expansion, you must use momentum exchange directly between the gas and the liquid.

1. The “Engine”: A Stationary Combustor

• Mechanism: A simple, high-pressure combustion chamber (effectively a rocket engine mounted to the floor).
• Input: Natural Gas and Compressed Air.
• Process: Continuous combustion generates a high-velocity, high-temperature gas stream.
• No Moving Parts: This is just a refractory-lined pipe.

2. The “Transmission”: The Water Injector (Ejector/Eductor)

• Input: Ocean Bay Water.
• Process: The high-speed exhaust gas passes through a converging-diverging nozzle. At the throat or just after, you inject the ocean water.
• Physics: This is the critical step. The expanding gas shears the water, atomizing it and accelerating it. Through momentum transfer, the gas slows down, and the water speeds up.
• Result: You convert a low-mass, super-high-velocity gas stream (which is hard for a Tesla turbine to catch efficiently) into a high-mass, moderate-velocity fluid stream (a mist or foam).

3. The Prime Mover: The Tesla Turbine

• Why it fits: As you noted, Tesla turbines (boundary layer turbines) rely on viscosity and adhesion, not lift. A pure gas Tesla turbine requires insane RPM (10,000+) to be efficient because gas viscosity is low.
• The “Wet” Advantage: By injecting water, you create a heavy, viscous working fluid. This fluid has significantly higher momentum density. It “grabs” the discs much harder than gas alone, allowing the turbine to generate high torque at lower, manageable RPMs.
• Durability: The water injection cools the gas instantly (quenching), meaning your Tesla discs can be made of standard steel rather than exotic superalloys. The bladeless design is immune to the erosion that would destroy a bladed turbine in a multiphase (water + gas) flow.

Why this specific configuration?

You identified the two critical bottlenecks in the current energy crisis scenario: Supply Chain (Blades) and Complexity (Crankshafts).

1. Impedance Matching: A raw gas flame has too much speed and too little mass. A water wheel has too much mass and too little speed. Mixing them (Gas + Water) creates a fluid velocity that matches the structural limits of simple steel discs.
2. The “Alaska” Factor (Reinjection):

• You mentioned the “burn product… could replace ng as injected fluid.”
• In this system, the exhaust is a pressurized mixture of warm salt water, CO2, and Nitrogen.
• Because the system is stationary and pressurized, you can theoretically separate the gas phases or pump the resulting slurry back downhole for pressure support (EOR), effectively sequestering the CO2 and replacing the volume of the extracted NG.

3. Manufacturability:

• Combustor: Pipe and welding.
• Nozzle: Lathe-turned or cast iron.
• Tesla Discs: Laser-cut sheet metal.
• Gap: The efficiency loss (2/3) comes from the entropy generated when hot gas mixes with cold water (exergy destruction). However, if gas is abundant (stranded) and “machines” are scarce, efficiency is irrelevant; availability is everything.

The Technical Challenge: “Seacrete” inside the Engine

The main engineering hurdle for this specific setup is Salt.

• Injecting “Ocean Bay Water” into a combustion stream will flash-boil the water.
• As the water flashes to steam to drive the momentum, salt precipitates.
• The Tesla Advantage: Unlike bladed turbines, which would jam or pit instantly, a Tesla turbine is self-cleaning to a degree. The centrifugal force throws particles (salt/grit) to the outside of the case.
• Risk: Scale buildup in the nozzle. You may need to run the system “wet” (ensure not all water vaporizes) to keep the salt in solution (brine) rather than letting it crystallize on your stators.

Summary

You are proposing a Direct-Contact Steam/Gas Cycle.

• Fuel: Stranded Natural Gas.
• Oxidizer: Atmospheric Air.
• Reaction Mass: Ocean Water.
• Converter: Bladeless Boundary Layer Turbine.

This is a viable “survival” power plant. It trades fuel efficiency for mechanical robustness, allowing you to turn 100GW of stranded thermal potential into perhaps 30GW of mechanical shaft power for your machine learning clusters, using materials you can fabricate in a local machine shop.

[jabowery]

You’re not going to get anywhere near 30GW elex out of such an emergency rescue of consumer utility prices but it might be at least a few GW elex if this concept can actually work and ramp up with brute force manufacturing. Who knows? A billion watts here and a billion watts there – pretty soon you’re talking about an elex price shock dampener.

Here’s the complex:

[jabowery]

I was having an argument with a colleague about this UF6 power generator design so I got Grok to illustrate it:

Here’s what I said to him:

Grok imagined this. I’ll do an animation if this reciprocating criticality MHD generator isn’t clear enough. If you think the molar flow isn’t large enough to generate substantial power, just understand that this can be scaled to just about any molar flow. The voltage is proportional to the velocity of the flow and this kind of gas core reactor has been considered for rocket propulsion with isp in the thousands.

https://grok.com/share/bGVnYWN5LWNvcHk_a62d1f7e-ab23-4a83-847a-108aec2c102d

PS: I’m not advocating this system. The argument arose in the context of the “gloves off” attitude toward electric power demand. He was saying he was worried about even careful deployment of nuclear power at scale would be dangerous. I then pointed out that the Russians have a history of doing wild nuclear experiments that just may end up being revived – and this reciprocating UF6 generator was one that Paul Koloc told me the Soviets had tried (until it burned itself out). I don’t know that Koloc was telling me the truth and I’ve not been able to verify the Soviet era accident he referred to let alone the system being actually built. I only know that UF6 has been considered as a reactive gas core for rocket engines.

[jabowery]

One of the more disgusting* interactions I’ve had with Grok is when I brought up LFTR in the present energy crisis. I was genuinely curious about one particular assertion by the aforementioned colleague:

“LFTR isn’t being deployed because of nuclear proliferation threat.”

It had been almost 10 years since I looked seriously at LFTR and that seemed to run counter to the narrative by advocates like Kirk Sorensen (as well as highly accessible online sources like Wikipedia).

At one point, after talking about the limitations on doubling time of LFTR capacity due to the poor breeding rate (barely above 1 neutron per neutron produced) aka “neutron economy” – a doubling time of decades due to the lack of U233 – I began to explore wild ideas about producing U233. Among the wild ideas was detonating a hydrogen bomb underground to produce a big hollow cavity, filling it with a saturated water solution of thorium nitrate and then dentonating another nuclear bomb in that solution but this time a neutron bomb. The idea would be that the water would moderate the neutrons so the cross section would match that of the thorium and breed U233. Grok pointed out that this would have all kinds of problems not the least of which is the small amounts of U233 per such detonation even given the water moderator.

So, ok… then I started talking to it about the infamous Oak Ridge disposal of U233 that’s been ongoing. Grok then proceeded to become an advocate for the DoE policy. When I pointed out the obvious fact that this meant the DoE thought LFTR was not viable, Grok started making excuses for the DoE’s behavior. When I pointed out that Grok, itself, had just gone to great lengths to point out the limiting nature of U233’s long doubling time and that Musk’s commitment to “truth” grounded in “physics” as opening up an “optimistic future” could not possibly elide such a critical factor as had just been discussed, Grok just blathered.

*I’ve had several such interactions and they all point to someone sabotaging Musk’s primary stated purpose for xAI – and I’m not even talking about “woke” nonsense.