SpaceX Is Making Its Own Suppliers OBSOLETE

Kind: captions Language: en SpaceX keeps moving into businesses that look less and less like rockets. In Florida, it is pursuing its own 32mm natural gas pipeline. And now Musk has confirmed talks with TSMC, the maker of the world's best chips. Meanwhile, the hardware for Flight 15 is already moving. And that flight helps decide whether any of this is even possible. In Louisiana, a fresh permit application just puts SpaceX's plans for giant roads and docks on paper. And SpaceX's own Starmine page lists numbers that make the TSMC story a lot bigger. So why is one rocket company touching natural gas chips, giant infrastructure, and AI computers all at the same time? To answer that, start with the strangest one, the gas. Right now, SpaceX gets its natural gas delivered by truck. SpaceX told the state's utility board that a single launch needs hundreds of truckloads of liqufied natural gas and that is only the gas. The liquid oxygen and all the other supplies come on top of it. Natural gas has to be chilled to around - 160° C to turn into a liquid. So every one of those loads is a rolling freezer on a public road. Picture that convoy for a second. Each truck has to be filled somewhere else, driven in and show up on time. That works for a rocket that flies now and then. For a rocket SpaceX wants flying constantly, it means a convoy of outside drivers, suppliers, and road conditions, all with a hand on SpaceX's stopwatch. So, SpaceX is building its way out. A SpaceX company called Coastal Connect Services wants to build a 32m gas pipeline 16 in wide. It would run entirely inside Breard County, the home of Cape Canaveral. It would tie into the Florida Gas Transmission System, the the big interstate line that carries gas into the state. And here is the detail that tells you who it is really for. According to the filing, the planned capacity is expected to be fully committed to SpaceX. This is a gas company built around exactly one customer. The obvious question is why SpaceX could not just plug into the gas lines that already exist. SpaceX checked. It looked at a pipeline run by Florida City Gas and told the state it fell short. It could not deliver to the right places in the right amounts at the right times or handle the big swings in how much gas a launch day needs. Upgrading that line would have come with terms, costs, and timing that did not meet SpaceX's requirements either. Plenty of companies will happily sell SpaceX methane. What the existing system cannot do is behave the way SpaceX wants its launch system to behave ready on SpaceX's schedule every single time. Look at what the pipe changes. Today, the chain runs from outside suppliers onto trucks, down public roads, and finally to the launch site. and SpaceX controls almost none of it. A traffic jam, a storm, or a driver shortage can all reach into the countdown. With the pipeline, gas flows from the big interstate line straight into a SpaceX system, and almost every link after that belongs to SpaceX. The paperwork is moving, too. Staff at Florida's Public Service Commission have recommended approving Coastal Connects rates, and the vote was put on the agenda for October 6th. That covers the business side. The pipe itself still needs its own building and environmental permits before any gas flows, and the plan goes a lot deeper than a pipe. SpaceX is also hiring for a job title you would never expect at a rocket company, Trader Natural Gas. It is based at Cape Canaveral, and according to the posting, this person will build and lead a brand new natural gas trading team. The job covers real gas deals and paper ones, managing bets in the market, and keeping an eye on pipelines, storage, prices, and even the weather. And that hire has company. SpaceX's career pages also list natural gas sourcing jobs, LNG project roles, and more gas trading positions at both Cape Canaveral and Starbase. That changes the whole story. A company that only wanted a pipe would hire pipe engineers. A company that builds a trading desk wants to understand the market itself, when to buy, how much to store, and how to protect itself when prices jump. Gas prices swing with the weather, with wars, and with winter demand. And a company burning this much gas can save or lose real money just by buying at the right time. Starship burns methane made from natural gas, so SpaceX will always need it. The smart move is to shrink the number of outside players who can turn a problem somewhere upstream into a delay on the launchpad. Big airports get most of their jet fuel through pipelines instead of trucks, and SpaceX wants its launch site to work the same way. Gas is the easy one, though, because lots of companies can sell you methane. The next thing SpaceX needs is one of the hardest things humans have ever learned to make. Elon Musk just confirmed that SpaceX has been talking with TSMC, the company that makes the world's most advanced chips. Reports said TSMC was looking at a role in a project called Terraab. Musk's answer was careful. In his words, just discussions, but something may come of it. So, keep this one firmly in the talks column. No deal has been announced. There is no confirmed ownership or chip process. And TSMC itself has said nothing about joining. Terrafab is the part that should make you sit up. It is a joint SpaceX and Tesla project and it exists because both companies believe their future need for computing will blow past what the world can supply. SpaceX's own announcement says their combined future demand is expected to top one terowatt of compute. So they want a chip operation of their own. It would cover the chips that do the thinking, the memory, the packaging, and the testing all in one connected system. And the plan is huge. It describes more than 100 million square ft, which is roughly the area of 1,700 American football fields. The first phase alone is estimated at about 16.8 billion for a project measured in neighborhoods instead of buildings. Credit where it is due though because chipmaking at TSMC's level took decades and plenty of rich companies and whole countries have spent fortunes trying to catch up. SpaceX is walking into the hardest factory business on the planet on purpose. And SpaceX already runs on chips more than most people realize. SpaceX is busy taking over industries. Help me take over the part of YouTube still pretending B-roll. Only videos are good. If 3D breakdowns are your thing, hit subscribe and support the channel. It genuinely means the world to me. Every Starlink satellite and every dish on a customer's roof is packed with them. And Starlink has millions of customers. Tesla has its own giant appetite for AI chips for its cars and its robots. Once your future runs on chips, somebody else gets to decide how many you get, when you get them, and what they cost. SpaceX says part of the Terraab output is meant for high power chips for its own data centers in space, which leaves two strange questions hanging at once. If this video is about SpaceX going to war with its suppliers, why is it sitting down with the most important chip supplier on Earth? And what could SpaceX possibly want to put in orbit that needs a chip project this big? Hold on to both because the answers are coming. And one of them is sitting on SpaceX's own website. But every chip in the world is useless if the rocket meant to carry it never becomes routine. And that rocket is about to fly again. That rocket is Starship and Flight 15 is starting to come together. People who track the hardware tie it to booster 22 and ship 42 and both finished their cryo tests in September. That is the test where a vehicle is filled with super cold liquid to prove its tanks can handle the pressure. Next spaceflight currently tracks the pair as the expected flight 15 hardware. So the rocket for the next test already exists and it has already passed its first big checkup. Then early on October 4th, RGV aerial photography caught a big test stand on the move. It is called the ship thrust simulator and it was heading from SpaceX's Massiey's test site back toward the factory. That stand is built to copy the push of a ship's engines during testing. The move itself is the fact. Some watchers read it as a sign that another ship test campaign is coming, and that part is their guess, separate from anything SpaceX has announced. And plenty of gates are still open. Engine work is the next step and both the ship and the booster still need static fires. The FAA still has to clear the flight. The rocket still has to be stacked on the pad and SpaceX has yet to publish a missions page or a launch window. One air traffic planning target has pointed to no earlier than October 19th, but trackers treat that as a rough plan, so it stays off the calendar for now. Then there is the question everybody wants answered. Flight 15 may be the first realistic chance for SpaceX to try catching the ship with the tower, the step that would make the whole rocket reusable. SpaceX has not listed that as an objective, so for now it is a possibility. But if it happens, the math behind everything in this video changes overnight. A booster that comes back is half of a reusable rocket. A ship that comes back and gets caught is the other half. And a fully reusable rocket is what makes all of these giant plans affordable. And this is where the title turns around and bites. Gas, trucks, pipelines, and chip shortages are all problems SpaceX can attack from the outside. Starship is a different kind of dependency because SpaceX is the supplier. The more SpaceX pulls inside the company, the more its future depends on SpaceX getting things right. Every giant plan SpaceX has quietly assumed Starship becomes repeatable instead of heroic and flight 15 is the next step in finding out which raises the obvious question. If SpaceX is preparing this much heavy industry before Starship has even matured, what scale does it think is coming? Louisiana gives a pretty clear hint. Flight 15 will show whether Starship can carry that kind of scale. And on the Gulf Coast, SpaceX is already preparing the ground for it with paperwork to prove it. For a while, SpaceX's plans there, a heavy hall road, a wararf on the Gulf, and a huge amount of digging, mostly lived in local reporting. Now, there is a named permit application behind them. On September 25th, an application for a coastal use permit was filed for a project with a very long name. It is called the Southwest Louisiana Coastal Infrastructure Improvement Project. The paperwork spells out the pieces. There is a heavy hall road of almost 16 miles, a raised bridge, a wararf on the Gulf, and about 5.2 million cubic yards of digging. Work is proposed to start in 2027. So, the real news is the paperwork itself. What used to be local chatter now has a permit number you can point to. Why lay freight routes this early? Because Starship is bigger than the world it has to move through. Even at Starbase, moving a booster a few miles down the road can mean closing that road while while the hardware crawls past. A booster is about 9 m wide, wider than two highway lanes, and far heavier than what normal roads and bridges are built to carry. At some point, a company has to choose between shrinking its machine to fit the world or building the world around its machine. SpaceX keeps choosing the second one. That is also why the wararf matters so much. A ship can carry loads that no road could ever handle. And for a rocket this big, the sea is the one highway with no overpasses. Think about how strange that looks next to its rivals. Most of them are still trying to get one new rocket flying from the pads they already have, while SpaceX is planning freight routes for launch sites that do not exist yet. To be precise about it, this permit does not approve any launch pads, towers, or launch operations. The talk of flights from Louisiana around 2029 and huge launch rates came from reporting on a community meeting and that is a separate story from this permit. So Louisiana shows the physical world SpaceX thinks it needs. Star Mind finally explains the demand and it closes the question we left hanging with TSMC. Star Mind is SpaceX's plan for computing in orbit and from a distance it can sound like a vague idea. SpaceX's own material is far more specific than that. Its official Starmine page lists a spacecraft called AI1. Fully deployed, it stands about 30 m tall with a wingspan of about 75 m, which is uh wider than the wings of a Boeing 747 on a machine built to live in orbit. And it is a computer first. SpaceX lists up to 250 kilowatts of compute at peak and about 175 kW on average, roughly what more than 100 American homes use. The computer parts are built to take chips from different suppliers and the results flow into Starlink over fast laser links. SpaceX also claims that doing this in orbit can cut the power spent on cooling by about 10 times compared with data centers on the ground. That matters because on the ground a new data center can wait years just to get hooked up to enough power. Then comes the line that answers the TSMC question. SpaceX says it is building a Gigasat factory in Bassrop, Texas to quickly build and launch thousands of AI satellites. By SpaceX's own target, that could start as soon as late 2027. Future satellites may not copy AI1 exactly, but the direction is clear. Thousands of flying computers need a mountain of chips every single year on a schedule nobody else controls. Compare that with Google, which is flying a small test of its own AI chips into orbit on a SpaceX ride share. Google is running an experiment while SpaceX is planning a factory. And SpaceX spells out the bottlenecks itself. On its own website, it names three things that limit AI in orbit, which are mass to orbit, power generation, and AI chips. That is why it points out that AI1 delivers about 75 kilowatts of compute for every ton it weighs because every ton has to be launched. Read that list of limits again because it is basically this whole video sitting on SpaceX's own page. Now put the whole stack together. The chips come from Terraab and outside foundaries like TSMC. The satellites come from the Gigasat factory and Starlink is the network that connects them. Every story in this video is one layer of the same machine. And notice which layer sits in the middle of all of it. Without a starship that flies often and cheaply, the chips have no ride. The satellites stay on the ground and the pipeline feeds an empty pad. So, here is what you have been watching. Natural Gas, TSMC, Flight 15, Louisiana, and Orbital AI looked completely unrelated at the start. And they all follow the same rule. Whenever something outside SpaceX gets big enough to limit its cost, its schedule, or how fast it can launch, SpaceX starts pulling that layer inside until nobody else can set SpaceX's pace again. Nobody important gets to hold the stopwatch. And SpaceX has played this game since the very beginning. In its early years, it found that buying rocket parts from the usual suppliers was slow and expensive, so it started making them itself. Gas and chips are simply the newest names on a very old list. Most companies manage their suppliers. SpaceX keeps a list of the ones it plans to stop needing. And the smart part is how selective it is. SpaceX's own filing says Terrafab is meant to add to its access to chips and that it expects to keep buying a significant share of its compute hardware from outside suppliers. That is why it can talk to TSMC and build Terraab at the same time. SpaceX is happy to have partners. What it refuses to have is a partner that becomes a choke point. And here is the twist that makes it even better. While SpaceX works to need everyone else less, the rest of the space world keeps finding new reasons to need SpaceX more. Amazon has bought Falcon 9 launches for its own satellite network. Google is flying its chip test on a SpaceX rocket and the Pentagon leans on SpaceX more every year. SpaceX keeps taking the stopwatch away from its suppliers while everyone else keeps handing theirs to SpaceX. And um because this channel is pro spacex with receipts, here is the risk. Every pipeline, fab, factory, and freight road only looks genius if Starship actually reaches the flight rate those investments assume. If that takes years longer than planned, all of it becomes very expensive infrastructure waiting for demand. That is the real bet underneath this whole video. SpaceX is building the industrial machine first and trusting its own rocket to catch up. That is what the title really means. SpaceX still has suppliers and partners and it will for a long time. But one by one, it is making sure that none of them can ever hold the stopwatch except SpaceX itself. So what do you think? Is SpaceX smart to take control of every bottleneck it can find, or is it picking too many fights at once?