Loading video...

Video Failed to Load

Go Home

Important numbers from Terafab announcement: More than $16.8 billion — first-phase investment from SpaceX and Tesla. 3,000+ new jobs — expected to be created. $30 million — Texas Enterprise Fund grant. 100 million square feet — planned Terafab size, around 9.3 million square meters. 10 million square feet —...

cb_doge's profile picture

DogeDesigner

1,896,307 subscribers

163,134 views • 1 day ago •via X (Twitter)

Science & TechnologyNews & Politics
LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

This gives you a sense of just how massive Terafab is going to be compared to Tesla Gigafactory Texas. It will be 10x the size (100 million sq ft) and output 1 TW per year. Insane scale.
0:23

This gives you a sense of just how massive Terafab is going to be compared to Tesla Gigafactory Texas. It will be 10x the size (100 million sq ft) and output 1 TW per year. Insane scale.

Nic Cruz Patane

44,776 views • 1 month ago

🚨 Elon Musk's reported Terafab project could be bigger than most people realize. 👀 Jo Bhakdi says the planned Tesla-SpaceX facility could reach 100 million square feet, about 10 times larger than Giga Texas, with a goal of producing around 1 terawatt of AI chips per year. He also highlighted Elon’s vision for AI in space, with plans to rapidly increase compute power in orbit over the next few years. If these plans become reality, SpaceX and Tesla may not just be building products. They could be building the infrastructure that powers the future of AI. $TSLA $SPCX
6:33

🚨 Elon Musk's reported Terafab project could be bigger than most people realize. 👀 Jo Bhakdi says the planned Tesla-SpaceX facility could reach 100 million square feet, about 10 times larger than Giga Texas, with a goal of producing around 1 terawatt of AI chips per year. He also highlighted Elon’s vision for AI in space, with plans to rapidly increase compute power in orbit over the next few years. If these plans become reality, SpaceX and Tesla may not just be building products. They could be building the infrastructure that powers the future of AI. $TSLA $SPCX

Herbert Ong

23,227 views • 1 month ago

TESLA AND SPACEX WILL INVEST $16.8B TO START BUILDING ‘TERAFAB’ CHIP FACTORY IN TEXAS Tesla and SpaceX said Thursday that “Terafab,” the advanced chip factory they are jointly developing, will be built in Grimes County, Texas, outside Houston, and that they will make an initial investment of $16.8 billion in the project. The factory will be “the largest and most valuable building on Earth by far,” Elon Musk, CEO of both companies, wrote in a post on Thursday. More than 100 million square feet of manufacturing space is planned for the facility, according to SpaceX. Full article:
0:53

TESLA AND SPACEX WILL INVEST $16.8B TO START BUILDING ‘TERAFAB’ CHIP FACTORY IN TEXAS Tesla and SpaceX said Thursday that “Terafab,” the advanced chip factory they are jointly developing, will be built in Grimes County, Texas, outside Houston, and that they will make an initial investment of $16.8 billion in the project. The factory will be “the largest and most valuable building on Earth by far,” Elon Musk, CEO of both companies, wrote in a post on Thursday. More than 100 million square feet of manufacturing space is planned for the facility, according to SpaceX. Full article:

FXHedge

19,248 views • 1 day ago

Terafab will be built in Grimes County, Texas In April, we broke ground on our research fab on the North Campus of Giga Texas – the precursor to Terafab. Both Tesla & SpaceX will need far more chips than current & future global production can supply. This is why we're building the largest chip manufacturing facility ever, with the goal of producing over 1 terawatt of compute per year The future is built in Texas
0:53

Terafab will be built in Grimes County, Texas In April, we broke ground on our research fab on the North Campus of Giga Texas – the precursor to Terafab. Both Tesla & SpaceX will need far more chips than current & future global production can supply. This is why we're building the largest chip manufacturing facility ever, with the goal of producing over 1 terawatt of compute per year The future is built in Texas

Tesla

5,035,318 views • 1 day ago

Does anyone realize that if SpaceX pulls this off, Terafab would become the largest building ever constructed by floor area? More than 100 million square feet of manufacturing space, over 5x the floor area of the current largest building on Earth. I wanted to wait before reporting because news was coming out quite rapidly after the until announcement. SpaceX and Tesla are planning to invest roughly $16.8 billion in the initial buildout in Grimes County, Texas. It will manufacture, package and test advanced logic and memory chips under one roof because their combined demand is expected to exceed 1 terawatt of compute, more than the current global supply. The buildout is already so massive that ASML, (the only supplier of the most advanced EUV lithography machines) is factoring Terafab demand into its own 2027 and 2028 capacity plans. Elon’s rough estimate is that 25% of its AI compute output will go toward Tesla Optimus and the other 75% toward AI spacecraft. ASML has to plan years in advance for the sheer amount of chipmaking capacity Elon wants for robots and ‘ai spacecraft’
0:53

Does anyone realize that if SpaceX pulls this off, Terafab would become the largest building ever constructed by floor area? More than 100 million square feet of manufacturing space, over 5x the floor area of the current largest building on Earth. I wanted to wait before reporting because news was coming out quite rapidly after the until announcement. SpaceX and Tesla are planning to invest roughly $16.8 billion in the initial buildout in Grimes County, Texas. It will manufacture, package and test advanced logic and memory chips under one roof because their combined demand is expected to exceed 1 terawatt of compute, more than the current global supply. The buildout is already so massive that ASML, (the only supplier of the most advanced EUV lithography machines) is factoring Terafab demand into its own 2027 and 2028 capacity plans. Elon’s rough estimate is that 25% of its AI compute output will go toward Tesla Optimus and the other 75% toward AI spacecraft. ASML has to plan years in advance for the sheer amount of chipmaking capacity Elon wants for robots and ‘ai spacecraft’

Chris

65,896 views • 1 day ago

tesla is merging with spaceX. this announcement did all but confirm it. a multi-trillion dollar space/ai/robotics GIANT, a moon settlement and at the heart of it - the Terafab - here's why this is nuts: - tesla plans to ship 1-10 BILLION robots PER year. problem: global chip capacity is shit, doesn't scale. - Tesla's 'Terafab' will produce 1 TERAwatt of compute. 80% of that will go to outer-space to train AI (grok) via spaceX - terafab will make 2 AI chips: 1 for humanoid robots/cars and 1 for AI datacenters in space. - by housing everything under 1 roof - chip production speeds up by an order of magnitude. unbeatable. - moon settlement: the sun doesn't set in space. solar power is 5X larger in low-orbit space. much more on the moon (24/7 too!) - so elon will build a moon settlement! there is ZERO chance tesla survives without spaceX transporting these ai chips. elon estimates training AI in space will be CHEAPER THAN ON EARTH in 2-3 years. insane. (heres the full announcement if you wanna watch)
21:45

tesla is merging with spaceX. this announcement did all but confirm it. a multi-trillion dollar space/ai/robotics GIANT, a moon settlement and at the heart of it - the Terafab - here's why this is nuts: - tesla plans to ship 1-10 BILLION robots PER year. problem: global chip capacity is shit, doesn't scale. - Tesla's 'Terafab' will produce 1 TERAwatt of compute. 80% of that will go to outer-space to train AI (grok) via spaceX - terafab will make 2 AI chips: 1 for humanoid robots/cars and 1 for AI datacenters in space. - by housing everything under 1 roof - chip production speeds up by an order of magnitude. unbeatable. - moon settlement: the sun doesn't set in space. solar power is 5X larger in low-orbit space. much more on the moon (24/7 too!) - so elon will build a moon settlement! there is ZERO chance tesla survives without spaceX transporting these ai chips. elon estimates training AI in space will be CHEAPER THAN ON EARTH in 2-3 years. insane. (heres the full announcement if you wanna watch)

Ejaaz

163,328 views • 4 months ago

NEWS: SpaceX has detailed Starmind, its plan for up to 1 million AI compute satellites in orbit, on a new website. Starmind moves AI data centers into space. Each satellite runs AI compute in orbit and beams results to Earth through Starlink lasers. The satellites draw constant solar power and cool for free in the vacuum. SpaceX filed that plan with the FCC in January. The first satellite, AI1, carries 150 kW of peak compute and spans 70 meters. Starship will carry 30 to 50 AI1 satellites per launch. SpaceX will build them at a new Gigasat factory in Bastrop, Texas. The AI chips come from a planned SpaceX and Tesla fab called Terafab. Two prototypes target early 2027, with thousands of satellites from late 2027.
0:51

NEWS: SpaceX has detailed Starmind, its plan for up to 1 million AI compute satellites in orbit, on a new website. Starmind moves AI data centers into space. Each satellite runs AI compute in orbit and beams results to Earth through Starlink lasers. The satellites draw constant solar power and cool for free in the vacuum. SpaceX filed that plan with the FCC in January. The first satellite, AI1, carries 150 kW of peak compute and spans 70 meters. Starship will carry 30 to 50 AI1 satellites per launch. SpaceX will build them at a new Gigasat factory in Bastrop, Texas. The AI chips come from a planned SpaceX and Tesla fab called Terafab. Two prototypes target early 2027, with thousands of satellites from late 2027.

Muskonomy

26,421 views • 28 days ago

Walter Isaacson: "As excited as we may be about LLMs these days, that's nothing compared to what happens when millions of Optimus robots and self-driving cars come." In Tesla's earnings report yesterday, they said that preparations for their first large-scale Optimus factory will begin in Q2. That first-generation line at Fremont is designed for 1 million robots a year. Tesla is also preparing Giga Texas for a second-generation line, which is being designed for long-term annual production capacity of 10 million robots.
0:49

Walter Isaacson: "As excited as we may be about LLMs these days, that's nothing compared to what happens when millions of Optimus robots and self-driving cars come." In Tesla's earnings report yesterday, they said that preparations for their first large-scale Optimus factory will begin in Q2. That first-generation line at Fremont is designed for 1 million robots a year. Tesla is also preparing Giga Texas for a second-generation line, which is being designed for long-term annual production capacity of 10 million robots.

Sawyer Merritt

136,500 views • 3 months ago

BREAKING: Elon Musk just announced the Terafab project >Tesla + SpaceX + xAI project >location: Austin/ Giga Texas >1TW/year of compute per year >logic + memory + advanced packaging in one fab >vertically integrated with recursive improvement >chips for FSD, Optimus, Grok, Dojo, Starlink >80% of compute going to space >20% for ground use >AI5: edge/inference for FSD robotaxi >AI6: powering Optimus >D3: space-optimized >runs HOTTER to minimize radiator mass >showed a 100kW AI Starlink Mini prototype >future AI Starlink will scale to megawatt range >Optimus: 1-10 BILLION units/year >100-200GW of compute for Optimus alone >Elon: "the next step towards becoming a galactic civilization" ITS HAPPENING
2:18

BREAKING: Elon Musk just announced the Terafab project >Tesla + SpaceX + xAI project >location: Austin/ Giga Texas >1TW/year of compute per year >logic + memory + advanced packaging in one fab >vertically integrated with recursive improvement >chips for FSD, Optimus, Grok, Dojo, Starlink >80% of compute going to space >20% for ground use >AI5: edge/inference for FSD robotaxi >AI6: powering Optimus >D3: space-optimized >runs HOTTER to minimize radiator mass >showed a 100kW AI Starlink Mini prototype >future AI Starlink will scale to megawatt range >Optimus: 1-10 BILLION units/year >100-200GW of compute for Optimus alone >Elon: "the next step towards becoming a galactic civilization" ITS HAPPENING

NIK

85,670 views • 4 months ago

Elon Musk announced a chip factory 10 times the size of Tesla's Gigafactory. The goal is to produce enough AI compute to equal twice the entire electricity consumption of the United States. He called it the Terafab. Here is the number that stopped me cold. The entire global AI chip industry right now is on track to hit around 100 gigawatts per year of compute. Every Nvidia GPU, every Google TPU, every chip from every company on earth combined. 100 gigawatts. Musk wants one factory to produce a terawatt per year. A terawatt is 1,000 gigawatts. Ten times the output of the entire global industry. From a single building. To put the scale in physical terms, the Terafab would need to be around 100 million square feet. You would need Starship point to point transport just to get from one end to the other. But the reason for the scale is not ambition for its own sake. To launch meaningful AI compute into space, you need a billion chips per year running at a kilowatt each. That is not a number the current industry can produce. The Terafab is the only way to get there. The timeline he put out: a gigawatt of space AI compute annualized by end of next year. Then 10x per year from there. 10 gigawatts by year two. 100 gigawatts by year three. A terawatt beyond that. Most people think orbital data centers are a decade away. Musk is building the factory to make them possible by next year.
3:26

Elon Musk announced a chip factory 10 times the size of Tesla's Gigafactory. The goal is to produce enough AI compute to equal twice the entire electricity consumption of the United States. He called it the Terafab. Here is the number that stopped me cold. The entire global AI chip industry right now is on track to hit around 100 gigawatts per year of compute. Every Nvidia GPU, every Google TPU, every chip from every company on earth combined. 100 gigawatts. Musk wants one factory to produce a terawatt per year. A terawatt is 1,000 gigawatts. Ten times the output of the entire global industry. From a single building. To put the scale in physical terms, the Terafab would need to be around 100 million square feet. You would need Starship point to point transport just to get from one end to the other. But the reason for the scale is not ambition for its own sake. To launch meaningful AI compute into space, you need a billion chips per year running at a kilowatt each. That is not a number the current industry can produce. The Terafab is the only way to get there. The timeline he put out: a gigawatt of space AI compute annualized by end of next year. Then 10x per year from there. 10 gigawatts by year two. 100 gigawatts by year three. A terawatt beyond that. Most people think orbital data centers are a decade away. Musk is building the factory to make them possible by next year.

Ihtesham Ali

316,910 views • 1 month ago

Elon Musk says Humanoid Robots will outnumber vehicle production - Vehicle production on Earth is about 100 million vehicles a year - Humanoid robot production will be between 1 billion and 10 billion units a year - Tesla will make the largest percentage of these robots Late 2026 mass production will ramp begins in earnest, with Optimus Gen 3/4 The long term goal will be for Optimus to become Tesla’s biggest product by volume and value Elon Musk says 80%+ of Tesla’s future value could come from robotics and AI
0:43

Elon Musk says Humanoid Robots will outnumber vehicle production - Vehicle production on Earth is about 100 million vehicles a year - Humanoid robot production will be between 1 billion and 10 billion units a year - Tesla will make the largest percentage of these robots Late 2026 mass production will ramp begins in earnest, with Optimus Gen 3/4 The long term goal will be for Optimus to become Tesla’s biggest product by volume and value Elon Musk says 80%+ of Tesla’s future value could come from robotics and AI

Wall Street Apes

66,366 views • 2 months ago

Bro… Elon just laid out the blueprint on how xAI and SpaceX are getting to 1,000+ gigawatts per year and beyond, in his closing statement at the xAI all-hands! IMO, this is why there will be no competition in space. 1/ Earth Supercomputers (Now) Build Memphis cluster to get us to >1 GW of power. “We’re only right now using roughly one percent of the potential energy of Earth.” Plan: • 330,000+ Grace Blackwell GPUs • ~1M H100-equivalent compute • 1 gigawatt draw at full scale • Built in <1 year • Tesla Megapacks stabilizing energy FYI, most AI companies are operating in hundreds of megawatts, yet xAI is already at a utility-scale gigawatt compute. 2/ Orbital Datacenters (Soon) Get to 100-200 GW per year launched into orbit with a path to ~1 terawatt (1,000 GW) total from Earth launches. “The next step beyond Earth data centers is our Earth orbital datacenters… launching at the 100-200 gigawatt per year level. Not cumulative, I mean per year.” Benefits: • Continuous solar exposure • No land constraints • No terrestrial grid bottlenecks • Virtually unlimited horizontal expansion FYI, 100–200 GW per year is equivalent to adding multiple large nation-scale grids annually. 3/ Moon Factories + Mass Driver (Mid Future) Get to 1,000+ GW per year, several orders of magnitude beyond Earth. “In order to do that you have to go to the Moon… We are actually going to have a mass driver on the Moon.” Plan: • Lunar factories build AI satellites • Electromagnetic mass driver launches them without fuel • Low lunar gravity reduces launch energy requirements • Scales far beyond Earth’s physical constraints FYI, this is when industrialized compute begins manufacturing off Earth. 4/ Solar System & Beyond (Long Term Future) Today the sun outputs ~3.8 × 10²⁶ watts “If we wanted to use even a millionth of the Sun’s energy, that would be roughly a million times more energy than civilization currently uses.” This means • 0.000001 of the Sun’s output = ~1,000,000 × today’s global Earth energy usage. 🤯 “Earth is really a tiny, tiny dust mote in a vast darkness… The Sun is 99.8% of all mass in the solar system.” To access that scale: • Moon manufacturing • Mars expansion • Solar-orbit compute clusters • Eventually tapping meaningful fractions of stellar output So… the blueprint to get here is 1/ Start at 1 GW. 2/ Scale to 1 TW. 3/ Scale to 1,000+ GW per year. 4/ Then expand toward fractions of the Sun itself. It’s clear that xAI + SpaceX is building the AI infrastructure and pathway to a stellar-scale energy civilization. I really hope I’m still alive to witness all this.
3:23

Bro… Elon just laid out the blueprint on how xAI and SpaceX are getting to 1,000+ gigawatts per year and beyond, in his closing statement at the xAI all-hands! IMO, this is why there will be no competition in space. 1/ Earth Supercomputers (Now) Build Memphis cluster to get us to >1 GW of power. “We’re only right now using roughly one percent of the potential energy of Earth.” Plan: • 330,000+ Grace Blackwell GPUs • ~1M H100-equivalent compute • 1 gigawatt draw at full scale • Built in <1 year • Tesla Megapacks stabilizing energy FYI, most AI companies are operating in hundreds of megawatts, yet xAI is already at a utility-scale gigawatt compute. 2/ Orbital Datacenters (Soon) Get to 100-200 GW per year launched into orbit with a path to ~1 terawatt (1,000 GW) total from Earth launches. “The next step beyond Earth data centers is our Earth orbital datacenters… launching at the 100-200 gigawatt per year level. Not cumulative, I mean per year.” Benefits: • Continuous solar exposure • No land constraints • No terrestrial grid bottlenecks • Virtually unlimited horizontal expansion FYI, 100–200 GW per year is equivalent to adding multiple large nation-scale grids annually. 3/ Moon Factories + Mass Driver (Mid Future) Get to 1,000+ GW per year, several orders of magnitude beyond Earth. “In order to do that you have to go to the Moon… We are actually going to have a mass driver on the Moon.” Plan: • Lunar factories build AI satellites • Electromagnetic mass driver launches them without fuel • Low lunar gravity reduces launch energy requirements • Scales far beyond Earth’s physical constraints FYI, this is when industrialized compute begins manufacturing off Earth. 4/ Solar System & Beyond (Long Term Future) Today the sun outputs ~3.8 × 10²⁶ watts “If we wanted to use even a millionth of the Sun’s energy, that would be roughly a million times more energy than civilization currently uses.” This means • 0.000001 of the Sun’s output = ~1,000,000 × today’s global Earth energy usage. 🤯 “Earth is really a tiny, tiny dust mote in a vast darkness… The Sun is 99.8% of all mass in the solar system.” To access that scale: • Moon manufacturing • Mars expansion • Solar-orbit compute clusters • Eventually tapping meaningful fractions of stellar output So… the blueprint to get here is 1/ Start at 1 GW. 2/ Scale to 1 TW. 3/ Scale to 1,000+ GW per year. 4/ Then expand toward fractions of the Sun itself. It’s clear that xAI + SpaceX is building the AI infrastructure and pathway to a stellar-scale energy civilization. I really hope I’m still alive to witness all this.

Teslaconomics

526,587 views • 5 months ago

Many still don’t understand why Elon is building Terafab Terafab is an extension to all the chip makers in the world It’s not about replacement, not a rivalry and absolutely not competing It’s being built to fulfill the massive chip orders that Tesla, SpaceX and xAI actually need TSMC’s most advanced 2nm capacity is totally booked through 2028 Tesla signed a massive $16.5 billion deal with Samsung back in July 2025 to produce AI6 chips at their Taylor, Texas factory and Samsung is building a Tesla Exclusive chip manufacturing plant to full fill this orders When Elon announced Terafab on March 21, 2026...he made it clear: “That rate is much less than we’d like. We either build the Terafab or we don’t have the chips, and we need the chips, so we build the Terafab” He basically told the chip makers: “Produce as much as you comfortably can. We will take them all. Actually we want even more” Even today Elon said: "SpaceX/Tesla will be always be major customers of TSMC and not competitors in the normal sense of the word" Current production rates are much less than they need....That’s why Terafab exists Terafab is an extension to every chip maker… not competition, not rivalry, absolutely not Even Intel has joined as a partner Even if chip supply improves, massive bottlenecks still exist with memory and advanced packaging You simply can't risk those supply chain breaks at this scale That’s why Terafab is being built to vertically integrate everything - chips, memory, advanced packaging all under one roof, targeting 1 terawatt of AI compute capacity per year This is a ludicrous amount of chips that no chipmaker currently produces at this scale. I don't think even TSMC and Samsung truly understand these numbers yet It’s about building the capacity the future actually demands
0:36

Many still don’t understand why Elon is building Terafab Terafab is an extension to all the chip makers in the world It’s not about replacement, not a rivalry and absolutely not competing It’s being built to fulfill the massive chip orders that Tesla, SpaceX and xAI actually need TSMC’s most advanced 2nm capacity is totally booked through 2028 Tesla signed a massive $16.5 billion deal with Samsung back in July 2025 to produce AI6 chips at their Taylor, Texas factory and Samsung is building a Tesla Exclusive chip manufacturing plant to full fill this orders When Elon announced Terafab on March 21, 2026...he made it clear: “That rate is much less than we’d like. We either build the Terafab or we don’t have the chips, and we need the chips, so we build the Terafab” He basically told the chip makers: “Produce as much as you comfortably can. We will take them all. Actually we want even more” Even today Elon said: "SpaceX/Tesla will be always be major customers of TSMC and not competitors in the normal sense of the word" Current production rates are much less than they need....That’s why Terafab exists Terafab is an extension to every chip maker… not competition, not rivalry, absolutely not Even Intel has joined as a partner Even if chip supply improves, massive bottlenecks still exist with memory and advanced packaging You simply can't risk those supply chain breaks at this scale That’s why Terafab is being built to vertically integrate everything - chips, memory, advanced packaging all under one roof, targeting 1 terawatt of AI compute capacity per year This is a ludicrous amount of chips that no chipmaker currently produces at this scale. I don't think even TSMC and Samsung truly understand these numbers yet It’s about building the capacity the future actually demands

X Freeze

53,291 views • 3 months ago

Everything Elon said about Optimus on the Q4 2024 earnings call: ⦿ I see a path for Tesla to be the most valuable company in the world, possibly bigger than the next five companies combined, overwhelmingly due to autonomous vehicles and autonomous humanoid robots. ⦿ The training compute needed for Optimus will ultimately probably be 10× what is needed for cars. Humanoids likely have 1,000× more useS than a car, which doesn't mean training scales by 1,000×, but probably close to 10×. The training compute will scale progressively as Optimus becomes more productive. ⦿ Long-term, Optimus has the potential to generate $10 trillion in revenue. In that scenario, we can support a lot of training compute. Even $500 billion in training compute is a good deal (chuckles). ⦿ There's a lot of uncertainty with timing because several aspects are being iterated simultaneously. The internal plan is for roughly 10,000 robots to be built this year, but we'll more likely produce several thousand. ⦿ I'm confident those several thousand robots will be able to do useful things. ⦿ The lessons from Production V1 will inform the changes in Production V2, which we expect to launch around mid-next year. ⦿ Our goal, aspirationally, is to ramp 10× every year, but perhaps we end up with 5× growth per year. With that kind of growth, it won't be many years before we're making 100 million robots a year. ⦿ The off-the-shelf components didn't work well, so we had to design everything in-house, including the most sophisticated hand ever made. Optimus will be able to play a piano and thread a needle. ⦿ My long-term prediction is that Optimus will overwhelmingly be the value of the company. ⦿ Optimus is not design-locked. It is rapidly evolving in a good direction. Tesla has by far the best humanoid robotics engineers in the world. Tesla also has all the other necessary ingredients: battery pack, power electronics, charging, communications, real-world AI, and the ability to scale production. ⦿ What other companies are missing is real-world AI and the ability to scale to millions of units a year. ⦿ This year, we aim to use Optimus internally at Tesla. We can easily use several thousand robots at Tesla for repetitive tasks, such as loading sheet metal at the welding line. ⦿ The Production V1 line is roughly 1,000 units per month. The Production V2, launching around mid-next year, will be for 10k units per month. The line after that will be for 100k units a month. Of course, it takes time for any given line to reach its maximum potential. ⦿ A very rough guess: we'll start delivering Optimus to companies outside of Tesla in the second half of 2026. The ramp is going to be exponential, and demand will not be a problem. ⦿ Once we're above 1 million units per year, the production cost of Optimus will be less than $20,000. Its total mass and complexity are much lower than a car. At a similar production volume to the Model Y, Optimus should be about half the cost of a Model Y. ⦿ The price is a different matter than cost. The price of Optimus will be set by market demand. [This is by far the longest Elon has ever spent discussing Optimus on an earnings call.]
12:28

Everything Elon said about Optimus on the Q4 2024 earnings call: ⦿ I see a path for Tesla to be the most valuable company in the world, possibly bigger than the next five companies combined, overwhelmingly due to autonomous vehicles and autonomous humanoid robots. ⦿ The training compute needed for Optimus will ultimately probably be 10× what is needed for cars. Humanoids likely have 1,000× more useS than a car, which doesn't mean training scales by 1,000×, but probably close to 10×. The training compute will scale progressively as Optimus becomes more productive. ⦿ Long-term, Optimus has the potential to generate $10 trillion in revenue. In that scenario, we can support a lot of training compute. Even $500 billion in training compute is a good deal (chuckles). ⦿ There's a lot of uncertainty with timing because several aspects are being iterated simultaneously. The internal plan is for roughly 10,000 robots to be built this year, but we'll more likely produce several thousand. ⦿ I'm confident those several thousand robots will be able to do useful things. ⦿ The lessons from Production V1 will inform the changes in Production V2, which we expect to launch around mid-next year. ⦿ Our goal, aspirationally, is to ramp 10× every year, but perhaps we end up with 5× growth per year. With that kind of growth, it won't be many years before we're making 100 million robots a year. ⦿ The off-the-shelf components didn't work well, so we had to design everything in-house, including the most sophisticated hand ever made. Optimus will be able to play a piano and thread a needle. ⦿ My long-term prediction is that Optimus will overwhelmingly be the value of the company. ⦿ Optimus is not design-locked. It is rapidly evolving in a good direction. Tesla has by far the best humanoid robotics engineers in the world. Tesla also has all the other necessary ingredients: battery pack, power electronics, charging, communications, real-world AI, and the ability to scale production. ⦿ What other companies are missing is real-world AI and the ability to scale to millions of units a year. ⦿ This year, we aim to use Optimus internally at Tesla. We can easily use several thousand robots at Tesla for repetitive tasks, such as loading sheet metal at the welding line. ⦿ The Production V1 line is roughly 1,000 units per month. The Production V2, launching around mid-next year, will be for 10k units per month. The line after that will be for 100k units a month. Of course, it takes time for any given line to reach its maximum potential. ⦿ A very rough guess: we'll start delivering Optimus to companies outside of Tesla in the second half of 2026. The ramp is going to be exponential, and demand will not be a problem. ⦿ Once we're above 1 million units per year, the production cost of Optimus will be less than $20,000. Its total mass and complexity are much lower than a car. At a similar production volume to the Model Y, Optimus should be about half the cost of a Model Y. ⦿ The price is a different matter than cost. The price of Optimus will be set by market demand. [This is by far the longest Elon has ever spent discussing Optimus on an earnings call.]

The Humanoid Hub

96,475 views • 1 year ago

Tesla is deploying $50 BILLION across 6 factories, a chip fab, robot production lines, AI supercomputers, lithium refineries, and solar manufacturing. To put that in perspective: Tesla made $477 million in profit last quarter. And is investing at roughly 100x that rate. Every other CEO on Earth would get fired for that ratio. Elon's doing it on purpose. Here's what he's assembling: - Own chip factory (TERAFAB with Intel, $25 billion, targeting 1 terawatt of AI compute per year) - Own energy grid (Megapacks powering entire cities) - Own robot workforce (Optimus production starting this year, 1 million units per year at Fremont, 10 million per year planned at Giga Texas) - Own transportation network (robotaxi live in Austin, Dallas, Houston with zero accidents, expanding to 9+ cities) - Own AI training infrastructure (Cortex 2 supercomputer online, 280,000 GPUs by June) - Own lithium refinery (Texas, ramping now) - Own solar panels (new design with 3x the power zones of conventional panels) - Own satellite compute (80% of TERAFAB output going to SpaceX orbital AI satellites) This is just insane. No company in history has attempted to own this many layers of its own supply chain simultaneously. Amazon took 20 years to become profitable because Bezos reinvested every dollar into infrastructure. Wall Street called him insane the entire time. Elon is running the same playbook but across MORE industries, at a FASTER pace, and with technology that didn't exist 5 years ago. The TERAFAB alone is designed to produce 70% of the output of the world's largest semiconductor foundry. Under one roof. Logic chips, memory, and packaging all vertically integrated. But why is he doing this? Elon said existing suppliers including TSMC, Samsung, and Micron simply cannot supply Tesla at the levels it needs. When you can't buy enough of what you need, you build the factory yourself. That's the Henry Ford playbook from 1920. Ford owned the rubber plantations, the iron mines, the glass factories, the railroads, and the forests that supplied his assembly lines. Elon is doing the same thing. Except his version includes orbital data centers, humanoid robots, and autonomous vehicles. The AI5 chip is already taped out. His team worked 6 months straight through holidays and weekends to finish early. He called it the best edge compute inference chip in existence. They're already designing AI6 AND Dojo 3. Meanwhile Tesla's FSD has 1.3 million paid subscribers globally. Record new subscriptions last quarter. Regulatory approval just landed in the Netherlands. China approvals expected by Q3. While every other automaker is trying to figure out how to compete with BYD on price, Elon is building the infrastructure layer that makes the car almost irrelevant. Because if you own the chips, the energy, the robots, the AI, the transportation network, AND the manufacturing... The car is just the interface. The real product is the ecosystem. Elon is spending $50 billion to build a parallel economy that doesn't depend on anyone else's supply chain, anyone else's chips, or anyone else's energy grid. That's closer to being a country than just a company. And whether you love him or hate him, nobody else alive is even attempting this.
1:26

Tesla is deploying $50 BILLION across 6 factories, a chip fab, robot production lines, AI supercomputers, lithium refineries, and solar manufacturing. To put that in perspective: Tesla made $477 million in profit last quarter. And is investing at roughly 100x that rate. Every other CEO on Earth would get fired for that ratio. Elon's doing it on purpose. Here's what he's assembling: - Own chip factory (TERAFAB with Intel, $25 billion, targeting 1 terawatt of AI compute per year) - Own energy grid (Megapacks powering entire cities) - Own robot workforce (Optimus production starting this year, 1 million units per year at Fremont, 10 million per year planned at Giga Texas) - Own transportation network (robotaxi live in Austin, Dallas, Houston with zero accidents, expanding to 9+ cities) - Own AI training infrastructure (Cortex 2 supercomputer online, 280,000 GPUs by June) - Own lithium refinery (Texas, ramping now) - Own solar panels (new design with 3x the power zones of conventional panels) - Own satellite compute (80% of TERAFAB output going to SpaceX orbital AI satellites) This is just insane. No company in history has attempted to own this many layers of its own supply chain simultaneously. Amazon took 20 years to become profitable because Bezos reinvested every dollar into infrastructure. Wall Street called him insane the entire time. Elon is running the same playbook but across MORE industries, at a FASTER pace, and with technology that didn't exist 5 years ago. The TERAFAB alone is designed to produce 70% of the output of the world's largest semiconductor foundry. Under one roof. Logic chips, memory, and packaging all vertically integrated. But why is he doing this? Elon said existing suppliers including TSMC, Samsung, and Micron simply cannot supply Tesla at the levels it needs. When you can't buy enough of what you need, you build the factory yourself. That's the Henry Ford playbook from 1920. Ford owned the rubber plantations, the iron mines, the glass factories, the railroads, and the forests that supplied his assembly lines. Elon is doing the same thing. Except his version includes orbital data centers, humanoid robots, and autonomous vehicles. The AI5 chip is already taped out. His team worked 6 months straight through holidays and weekends to finish early. He called it the best edge compute inference chip in existence. They're already designing AI6 AND Dojo 3. Meanwhile Tesla's FSD has 1.3 million paid subscribers globally. Record new subscriptions last quarter. Regulatory approval just landed in the Netherlands. China approvals expected by Q3. While every other automaker is trying to figure out how to compete with BYD on price, Elon is building the infrastructure layer that makes the car almost irrelevant. Because if you own the chips, the energy, the robots, the AI, the transportation network, AND the manufacturing... The car is just the interface. The real product is the ecosystem. Elon is spending $50 billion to build a parallel economy that doesn't depend on anyone else's supply chain, anyone else's chips, or anyone else's energy grid. That's closer to being a country than just a company. And whether you love him or hate him, nobody else alive is even attempting this.

Ricardo

138,597 views • 3 months ago

elon musk just sat down with his SpaceX team and explained exactly how they're going to turn humanity into a kardashev type 2 civilization it's one of the most insane things i've ever watched: 1. humanity currently uses less than one trillionth of the sun's energy output. a trillion is a million times a million. on the kardashev scale, the one physicists use to measure how advanced a civilization actually is, we are not even registering. we are effectively non-existent. 2. starship is the first rocket in history designed to be fully reusable. every other mode of transport, cars, planes, ships, bicycles, you take reusability for granted. rockets have always been thrown away after one use. if you had to throw away the plane after every flight, almost nobody would be flying. 3. spacex currently launches 85 to 90% of all mass to orbit on earth. the rest of the world, including most of the us, accounts for maybe 5 to 7%. that's before starship even gets going. 4. the plan is to go from 2,500 tons to orbit per year to a million tons per year. in roughly 3 years. that's not a projection. that's the internal target. 5. data centers are moving to space. by end of next year spacex is targeting 1 gigawatt of ai compute in orbit. then 10x every year after that. 10 gigawatts in 2.5 years. 100 gigawatts in 3.5 years. a terawatt eventually, which is twice the entire electricity consumption of the united states. 6. the ai satellite is actually simpler to build than a starlink satellite. it's mostly solar panels, a radiator, and a rack of gpus. the hard part was already solved building starlink. they're just making it bigger. 7. latency from orbit is about 3 milliseconds. light travels 300 km per millisecond. some people assume orbital compute means high latency. it doesn't. it's 3 milliseconds away. 8. the terafab will be 100 million square feet. ten times the size of the tesla gigafactory texas. the entire global chip industry is on track to hit maybe 100 gigawatts of ai compute per year. a terawatt requires a completely different order of manufacturing. that's why they're building it themselves. 9. to go beyond a terawatt you have to go to the moon. no atmosphere. one-sixth earth's gravity. you manufacture solar panels and radiators directly from moon materials. then you launch ai satellites into deep space using an electromagnetic rail gun. no rocket needed. musk calls this the mass driver. this is the next step on the actual roadmap. 10. if enough mass is going to the moon to run a rail gun operation at that scale, it also means regular people can go. musk's exact words: "i think everyone should go to the moon at least once." 11. the ai satellite has about a terabit of laser link connectivity. it connects to the starlink constellation which then sends data to the ground using frequencies that penetrate clouds and even roofs. the connection never drops regardless of weather. 12. the reference design for the first ai satellites is built around nvidia rubin and GB300 chips. but the architecture is open. google TPUs, amazon trainium, any chip can go up. spacex is building the infrastructure, not locking in the compute. 13. spacex is the only operator on earth with experience running a constellation at the scale of 10,000 satellites. nobody else is even close. that operational knowledge is a moat that cannot be replicated quickly.
31:08

elon musk just sat down with his SpaceX team and explained exactly how they're going to turn humanity into a kardashev type 2 civilization it's one of the most insane things i've ever watched: 1. humanity currently uses less than one trillionth of the sun's energy output. a trillion is a million times a million. on the kardashev scale, the one physicists use to measure how advanced a civilization actually is, we are not even registering. we are effectively non-existent. 2. starship is the first rocket in history designed to be fully reusable. every other mode of transport, cars, planes, ships, bicycles, you take reusability for granted. rockets have always been thrown away after one use. if you had to throw away the plane after every flight, almost nobody would be flying. 3. spacex currently launches 85 to 90% of all mass to orbit on earth. the rest of the world, including most of the us, accounts for maybe 5 to 7%. that's before starship even gets going. 4. the plan is to go from 2,500 tons to orbit per year to a million tons per year. in roughly 3 years. that's not a projection. that's the internal target. 5. data centers are moving to space. by end of next year spacex is targeting 1 gigawatt of ai compute in orbit. then 10x every year after that. 10 gigawatts in 2.5 years. 100 gigawatts in 3.5 years. a terawatt eventually, which is twice the entire electricity consumption of the united states. 6. the ai satellite is actually simpler to build than a starlink satellite. it's mostly solar panels, a radiator, and a rack of gpus. the hard part was already solved building starlink. they're just making it bigger. 7. latency from orbit is about 3 milliseconds. light travels 300 km per millisecond. some people assume orbital compute means high latency. it doesn't. it's 3 milliseconds away. 8. the terafab will be 100 million square feet. ten times the size of the tesla gigafactory texas. the entire global chip industry is on track to hit maybe 100 gigawatts of ai compute per year. a terawatt requires a completely different order of manufacturing. that's why they're building it themselves. 9. to go beyond a terawatt you have to go to the moon. no atmosphere. one-sixth earth's gravity. you manufacture solar panels and radiators directly from moon materials. then you launch ai satellites into deep space using an electromagnetic rail gun. no rocket needed. musk calls this the mass driver. this is the next step on the actual roadmap. 10. if enough mass is going to the moon to run a rail gun operation at that scale, it also means regular people can go. musk's exact words: "i think everyone should go to the moon at least once." 11. the ai satellite has about a terabit of laser link connectivity. it connects to the starlink constellation which then sends data to the ground using frequencies that penetrate clouds and even roofs. the connection never drops regardless of weather. 12. the reference design for the first ai satellites is built around nvidia rubin and GB300 chips. but the architecture is open. google TPUs, amazon trainium, any chip can go up. spacex is building the infrastructure, not locking in the compute. 13. spacex is the only operator on earth with experience running a constellation at the scale of 10,000 satellites. nobody else is even close. that operational knowledge is a moat that cannot be replicated quickly.

Jaynit

78,177 views • 1 month ago

Elon Musk just described a project so large that most people will assume he is exaggerating (Save this). He is not. In the video, Musk lays out the central problem facing every AI company on earth, the entire global chip industry is on a path to produce roughly 100 gigawatts of AI compute per year. That sounds like a lot until you understand that his companies alone Tesla, SpaceX, and xAI will need orders of magnitude more than that. His answer is the TerraFab. It is a joint chip factory spanning 100 million square feet, ten times the size of Tesla's Gigafactory Texas announced in March 2026, with Grimes County, Texas commissioners approving the full scale facility site just last week. The goal is one full terawatt of AI compute output per year. For context, 1 terawatt is 1,000 gigawatts twice the current total electricity consumption of the United States. SpaceX has already committed an initial $55 billion to the prototype phase, with total investment estimates ranging into the trillions. Here is why this matters for Micron specifically. In the video, Musk named Nvidia's Rubin chips as the reference design for TerraFab's first orbital deployments, and said "You're going to need a lot of memory to go with that." A billion full radical equivalent chips per year, each requiring stacks of high bandwidth memory, that is the demand signal Micron just received from one of the most capital-intensive projects in human history. And Micron already cannot keep up with what exists today. Micron's entire 2026 HBM output is fully sold out contracted before the year began. HBM4 entered volume production ahead of schedule and sold out immediately. The structural reason Micron wins here is simple. Every AI chip ever built Nvidia H100s, Rubin chips, custom ASICs, TPUs is useless without high-bandwidth memory stacked directly on top of it. There are only three companies in the world that supply HBM at scale, Samsung, SK Hynix, and Micron. Samsung has had quality issues, SK Hynix is supply constrained. Micron is the only US headquartered HBM manufacturer which matters enormously given CHIPS Act subsidies, domestic procurement requirements, and the political push to keep critical AI memory production on American soil. TerraFab just made the memory deficit permanently larger. Come join Milk Road Pro for our full breakdown of Micron and our entire AI thesis just for $1. Link below!
3:30

Elon Musk just described a project so large that most people will assume he is exaggerating (Save this). He is not. In the video, Musk lays out the central problem facing every AI company on earth, the entire global chip industry is on a path to produce roughly 100 gigawatts of AI compute per year. That sounds like a lot until you understand that his companies alone Tesla, SpaceX, and xAI will need orders of magnitude more than that. His answer is the TerraFab. It is a joint chip factory spanning 100 million square feet, ten times the size of Tesla's Gigafactory Texas announced in March 2026, with Grimes County, Texas commissioners approving the full scale facility site just last week. The goal is one full terawatt of AI compute output per year. For context, 1 terawatt is 1,000 gigawatts twice the current total electricity consumption of the United States. SpaceX has already committed an initial $55 billion to the prototype phase, with total investment estimates ranging into the trillions. Here is why this matters for Micron specifically. In the video, Musk named Nvidia's Rubin chips as the reference design for TerraFab's first orbital deployments, and said "You're going to need a lot of memory to go with that." A billion full radical equivalent chips per year, each requiring stacks of high bandwidth memory, that is the demand signal Micron just received from one of the most capital-intensive projects in human history. And Micron already cannot keep up with what exists today. Micron's entire 2026 HBM output is fully sold out contracted before the year began. HBM4 entered volume production ahead of schedule and sold out immediately. The structural reason Micron wins here is simple. Every AI chip ever built Nvidia H100s, Rubin chips, custom ASICs, TPUs is useless without high-bandwidth memory stacked directly on top of it. There are only three companies in the world that supply HBM at scale, Samsung, SK Hynix, and Micron. Samsung has had quality issues, SK Hynix is supply constrained. Micron is the only US headquartered HBM manufacturer which matters enormously given CHIPS Act subsidies, domestic procurement requirements, and the political push to keep critical AI memory production on American soil. TerraFab just made the memory deficit permanently larger. Come join Milk Road Pro for our full breakdown of Micron and our entire AI thesis just for $1. Link below!

Milk Road AI

246,975 views • 1 month ago

"Optimus will be the biggest product ever." - Elon Musk AND THIS IS HOW HE'S GOING TO DO IT.... REDEFINING LABOR ON EARTH The core mission of Optimus is to take over dangerous, repetitive, and monotonous work, making advanced automation accessible to everyone and fundamentally changing what human labor looks like. As a first step, Tesla plans to deploy 1,000 Optimus robots inside the cleanrooms of its new Terafab chip facility by 2027, with the goal of cutting human manufacturing errors by 90%. PRODUCTION TIMELINES & PRICING Musk has outlined a clear ramp: limited production is targeted for 2025, starting with more than 1,000 units for internal use across Tesla factories. External sales to other companies could begin in 2026. By 2027, Tesla aims to start high-volume production at Gigafactory Texas with a target of 10 million robots per year. Ultimately, Musk wants to bring the consumer price down to roughly $20,000–$30,000. "GEN 3" DEXTERITY UPGRADES The robot's hardware is advancing quickly. Tesla recently introduced the Gen 3 hands, which feature 22 degrees of freedom and 50 total actuators — roughly doubling the dexterity of earlier versions. These new hands include tactile fingertip sensors with force-torque feedback, enabling Optimus to handle delicate items such as eggs or glass vials without damaging them. THE "DIGITAL OPTIMUS" BRAIN To give the robot both physical skill and real intelligence, Tesla and xAI are developing an architecture called "Digital Optimus" (sometimes referred to internally as Macrohard). It uses a dual-process system: Tesla's vision-based driving AI serves as "System 1," handling fast, instinctive physical actions and balance. xAI's Grok model acts as "System 2," managing high-level reasoning, task planning, and natural conversation. NEXT-GENERATION SILICON Running powerful AI models on a mobile robot requires extreme efficiency. Optimus will be powered by Tesla's upcoming AI5 and AI6 edge-inference chips (produced at the Terafab facility), delivering 40 to 50 times the compute performance of today's chips while keeping power consumption low enough for all-day operation. LUNAR COLONIZATION AND SPACE OPERATIONS While earlier plans called for sending Optimus robots to Mars in 2026, SpaceX has shifted initial focus toward establishing a Moon colony first. Under Musk's "Kardashev Blueprint," large fleets of Optimus robots — referred to as "Optimi" — will be deployed to the lunar surface. Working autonomously around the clock, they will construct habitats, manufacture space-based AI satellites, and build a massive electromagnetic mass driver to launch the "Starmind" AI constellation into deep space.
1:15

"Optimus will be the biggest product ever." - Elon Musk AND THIS IS HOW HE'S GOING TO DO IT.... REDEFINING LABOR ON EARTH The core mission of Optimus is to take over dangerous, repetitive, and monotonous work, making advanced automation accessible to everyone and fundamentally changing what human labor looks like. As a first step, Tesla plans to deploy 1,000 Optimus robots inside the cleanrooms of its new Terafab chip facility by 2027, with the goal of cutting human manufacturing errors by 90%. PRODUCTION TIMELINES & PRICING Musk has outlined a clear ramp: limited production is targeted for 2025, starting with more than 1,000 units for internal use across Tesla factories. External sales to other companies could begin in 2026. By 2027, Tesla aims to start high-volume production at Gigafactory Texas with a target of 10 million robots per year. Ultimately, Musk wants to bring the consumer price down to roughly $20,000–$30,000. "GEN 3" DEXTERITY UPGRADES The robot's hardware is advancing quickly. Tesla recently introduced the Gen 3 hands, which feature 22 degrees of freedom and 50 total actuators — roughly doubling the dexterity of earlier versions. These new hands include tactile fingertip sensors with force-torque feedback, enabling Optimus to handle delicate items such as eggs or glass vials without damaging them. THE "DIGITAL OPTIMUS" BRAIN To give the robot both physical skill and real intelligence, Tesla and xAI are developing an architecture called "Digital Optimus" (sometimes referred to internally as Macrohard). It uses a dual-process system: Tesla's vision-based driving AI serves as "System 1," handling fast, instinctive physical actions and balance. xAI's Grok model acts as "System 2," managing high-level reasoning, task planning, and natural conversation. NEXT-GENERATION SILICON Running powerful AI models on a mobile robot requires extreme efficiency. Optimus will be powered by Tesla's upcoming AI5 and AI6 edge-inference chips (produced at the Terafab facility), delivering 40 to 50 times the compute performance of today's chips while keeping power consumption low enough for all-day operation. LUNAR COLONIZATION AND SPACE OPERATIONS While earlier plans called for sending Optimus robots to Mars in 2026, SpaceX has shifted initial focus toward establishing a Moon colony first. Under Musk's "Kardashev Blueprint," large fleets of Optimus robots — referred to as "Optimi" — will be deployed to the lunar surface. Working autonomously around the clock, they will construct habitats, manufacture space-based AI satellites, and build a massive electromagnetic mass driver to launch the "Starmind" AI constellation into deep space.

Lacey

68,739 views • 1 month ago

The sun produces more energy in one second than humanity has used in its entire existence. SpaceX thinks the only way to actually use more of it is to leave Earth. Here's why Elon Musk is racing to build data centers in space: 1. Musk frames civilizational progress using the Kardashev scale, a measurement created by a Russian physicist that ranks civilizations by how much energy they can harness. Type one means harnessing a planet's full energy, type two means harnessing a star's energy, and type three means harnessing a galaxy's energy. 2. Right now, humanity registers as essentially nonexistent on this scale. We harness less than a trillionth of the sun's total power output. Musk says we are not even at the level of a micro soul on the scale. 3. The sun makes up 99.86 percent of all mass in the entire solar system. Earth is so small in comparison that it falls into the leftover miscellaneous category alongside everything that is not the sun or Jupiter. 4. Only about half a billionth of the sun's energy even reaches Earth's cross section, and most of that cannot be used because 70 percent of Earth is covered in water, and much of the remaining land is uninhabitable terrain like Antarctica and Siberia. 5. To meaningfully climb the Kardashev scale, humanity has to go to space. Reaching even one millionth of the sun's total energy output would require increasing civilizational energy use by more than a million times current levels. 6. Getting to just 1 percent of the sun's energy would make a civilization vastly more powerful than ours is today. Musk says even reaching that level would represent an extremely advanced civilization. 7. Three core requirements stand between humanity and this goal: mass to orbit capability, a massive amount of solar power, and enough AI chips to actually use that power. 8. Starship solves the mass to orbit problem. It is designed to be the first rocket in history with full and rapid reusability, a breakthrough Musk calls absolutely necessary for making life multi-planetary and for ascending the Kardashev scale at all. 9. Reusability is the same principle behind every successful mode of transport. Cars, planes, boats, and bicycles are all reusable. If airplanes were thrown away after every flight, flying would be far too expensive for anyone to use. 10. Starship is already the largest, heaviest, and most powerful flying object ever built. Version 3 produces more than double the thrust of the Saturn V moon rocket, and version 4 is expected to produce nearly three times that thrust. 11. SpaceX currently delivers between 85 and 90 percent of all mass that reaches orbit from Earth using Falcon 9 and Falcon Heavy. With Starship, the company aims to scale mass to orbit from roughly 2,500 tons a year to millions of tons a year within about three years. 12. The proposed AI satellites are actually simpler to build than Starlink satellites. They mainly need solar cells, radiators, and laser links, without the complex phased array and parabolic antennas that Starlink satellites require. 13. The first version of the SpaceX AI satellite targets 150 kilowatts of peak power and 120 kilowatts of sustained power, roughly matching the output of a single Nvidia GB300 compute rack here on Earth. 14. These satellites will connect to each other through laser links and to the Starlink constellation, which then relays data to the ground. Despite orbiting 600 to 800 kilometers above Earth, the added latency is roughly only 3 milliseconds. 15. Heat management in space is actually easier than on Earth because radiators can simply release heat directly into the vacuum, removing the need for the massive cooling infrastructure required by ground based data centers. 16. SpaceX already operates around 10,000 Starlink satellites and claims to be the only company with real experience safely operating constellations at that scale, which gives them a head start in managing potentially thousands or even up to a million AI satellites. 17. To actually scale chip production to the levels needed, SpaceX is planning what it calls a Terafab, a chip manufacturing facility expected to span roughly 100 million square feet, about ten times the size of the existing Tesla Gigafactory in Texas. 18. The rough timeline targets reaching an annualized rate of 1 gigawatt of space based AI compute by the end of next year, scaling by roughly 10x per year afterward, eventually aiming for a terawatt per year, which is twice the entire current electricity consumption of the United States. 19. To push three orders of magnitude beyond even that terawatt goal, Musk describes building a mass driver on the moon, an electromagnetic rail gun style system that uses the moon's lack of atmosphere and lower gravity to launch satellites into space without needing a rocket at all. 20. Musk frames the long term vision in deeply personal terms. If enough mass and infrastructure eventually moves to the moon, it would become accessible enough that almost anyone who wants to go could go, and potentially even live there permanently. Follow Brad if you want more content on business, mindset & life changing ideas.
30:32

The sun produces more energy in one second than humanity has used in its entire existence. SpaceX thinks the only way to actually use more of it is to leave Earth. Here's why Elon Musk is racing to build data centers in space: 1. Musk frames civilizational progress using the Kardashev scale, a measurement created by a Russian physicist that ranks civilizations by how much energy they can harness. Type one means harnessing a planet's full energy, type two means harnessing a star's energy, and type three means harnessing a galaxy's energy. 2. Right now, humanity registers as essentially nonexistent on this scale. We harness less than a trillionth of the sun's total power output. Musk says we are not even at the level of a micro soul on the scale. 3. The sun makes up 99.86 percent of all mass in the entire solar system. Earth is so small in comparison that it falls into the leftover miscellaneous category alongside everything that is not the sun or Jupiter. 4. Only about half a billionth of the sun's energy even reaches Earth's cross section, and most of that cannot be used because 70 percent of Earth is covered in water, and much of the remaining land is uninhabitable terrain like Antarctica and Siberia. 5. To meaningfully climb the Kardashev scale, humanity has to go to space. Reaching even one millionth of the sun's total energy output would require increasing civilizational energy use by more than a million times current levels. 6. Getting to just 1 percent of the sun's energy would make a civilization vastly more powerful than ours is today. Musk says even reaching that level would represent an extremely advanced civilization. 7. Three core requirements stand between humanity and this goal: mass to orbit capability, a massive amount of solar power, and enough AI chips to actually use that power. 8. Starship solves the mass to orbit problem. It is designed to be the first rocket in history with full and rapid reusability, a breakthrough Musk calls absolutely necessary for making life multi-planetary and for ascending the Kardashev scale at all. 9. Reusability is the same principle behind every successful mode of transport. Cars, planes, boats, and bicycles are all reusable. If airplanes were thrown away after every flight, flying would be far too expensive for anyone to use. 10. Starship is already the largest, heaviest, and most powerful flying object ever built. Version 3 produces more than double the thrust of the Saturn V moon rocket, and version 4 is expected to produce nearly three times that thrust. 11. SpaceX currently delivers between 85 and 90 percent of all mass that reaches orbit from Earth using Falcon 9 and Falcon Heavy. With Starship, the company aims to scale mass to orbit from roughly 2,500 tons a year to millions of tons a year within about three years. 12. The proposed AI satellites are actually simpler to build than Starlink satellites. They mainly need solar cells, radiators, and laser links, without the complex phased array and parabolic antennas that Starlink satellites require. 13. The first version of the SpaceX AI satellite targets 150 kilowatts of peak power and 120 kilowatts of sustained power, roughly matching the output of a single Nvidia GB300 compute rack here on Earth. 14. These satellites will connect to each other through laser links and to the Starlink constellation, which then relays data to the ground. Despite orbiting 600 to 800 kilometers above Earth, the added latency is roughly only 3 milliseconds. 15. Heat management in space is actually easier than on Earth because radiators can simply release heat directly into the vacuum, removing the need for the massive cooling infrastructure required by ground based data centers. 16. SpaceX already operates around 10,000 Starlink satellites and claims to be the only company with real experience safely operating constellations at that scale, which gives them a head start in managing potentially thousands or even up to a million AI satellites. 17. To actually scale chip production to the levels needed, SpaceX is planning what it calls a Terafab, a chip manufacturing facility expected to span roughly 100 million square feet, about ten times the size of the existing Tesla Gigafactory in Texas. 18. The rough timeline targets reaching an annualized rate of 1 gigawatt of space based AI compute by the end of next year, scaling by roughly 10x per year afterward, eventually aiming for a terawatt per year, which is twice the entire current electricity consumption of the United States. 19. To push three orders of magnitude beyond even that terawatt goal, Musk describes building a mass driver on the moon, an electromagnetic rail gun style system that uses the moon's lack of atmosphere and lower gravity to launch satellites into space without needing a rocket at all. 20. Musk frames the long term vision in deeply personal terms. If enough mass and infrastructure eventually moves to the moon, it would become accessible enough that almost anyone who wants to go could go, and potentially even live there permanently. Follow Brad if you want more content on business, mindset & life changing ideas.

Brad

12,955 views • 1 month ago

Starship can reach Mars. But stainless steel mass makes return impossible. And now TeraFab’s orbital AI compute is becoming the enabling layer that closes the gap. The moment you attempt a crewed round-trip, you need: • 2,400 metric tons of ISRU propellant on the Martian surface • Structural mass trimmed by microns across millions of design permutations • Radiation-hardened autonomy that survives years in deep space • First-wave construction before any humans arrive • Launch cadence that collapses cost to ~$100/kg TeraFab is evolving into the silicon backbone for true multiplanetary capability. - $119B Texas facility producing 1 terawatt of custom silicon per year - 80% radiation-hardened D3 chips powering orbital simulation clusters - Millions of Starship variants digitally flown in weeks - 20% AI5 edge chips for Optimus legions that build propellant plants - 135 Starship flights/day delivering orbital compute - Revenue loop that self-funds the Mars fleet This is much bigger than “just a chip factory” It’s the bridge that turns an engineering impossibility into a scheduled logistics problem and moves manufacturing off Earth’s gravity well forever.
5:27

Starship can reach Mars. But stainless steel mass makes return impossible. And now TeraFab’s orbital AI compute is becoming the enabling layer that closes the gap. The moment you attempt a crewed round-trip, you need: • 2,400 metric tons of ISRU propellant on the Martian surface • Structural mass trimmed by microns across millions of design permutations • Radiation-hardened autonomy that survives years in deep space • First-wave construction before any humans arrive • Launch cadence that collapses cost to ~$100/kg TeraFab is evolving into the silicon backbone for true multiplanetary capability. - $119B Texas facility producing 1 terawatt of custom silicon per year - 80% radiation-hardened D3 chips powering orbital simulation clusters - Millions of Starship variants digitally flown in weeks - 20% AI5 edge chips for Optimus legions that build propellant plants - 135 Starship flights/day delivering orbital compute - Revenue loop that self-funds the Mars fleet This is much bigger than “just a chip factory” It’s the bridge that turns an engineering impossibility into a scheduled logistics problem and moves manufacturing off Earth’s gravity well forever.

Lacey

49,029 views • 2 months ago