正在加载视频...

视频加载失败

返回首页

Goldman Sachs just quintupled its robot forecast and Morgan Stanley thinks even that is far too conservative (Save this). Goldman's old base case called for 1.4 million humanoid units in 2035 and the new one calls for 6.48 million. The 2030 estimate jumped from 256,000 to 890,000 units, putting...

MelvinInvests's profile picture

Melvin

24,961 subscribers

31,961 次观看 • 9 天前 •via X (Twitter)

科学技术财经新闻政治
LIVE

Anya Rossi• Live Now

Private livecam show

19 条评论

Micro2Macr0 的头像
Micro2Macr09 天前

So self. Driving cars count? If so, it'll be that number by next year. 😂

Melvin 的头像
Melvin8 天前

I have a feeling Elon will be coming in hot next year with Cybercabs

Kyle Reidhead | Milk Road 的头像
Kyle Reidhead | Milk Road9 天前

ChatGPT moment for robotics is coming very soon

Melvin 的头像
Melvin9 天前

coming near you very soon

Тигран Казарян 的头像
Тигран Казарян9 天前

Experts often evaluate data science by measuring real-world performance rather than theoretical potential in large-scale production environments.

Ace-Kicker 的头像
Ace-Kicker9 天前

Love this update — robot supply chain is the underappreciated AI tape. Actuators, reducers, and power electronics look like the next semis-style bottleneck once unit forecasts keep jumping. Sharp save.

Yasser Khan 的头像
Yasser Khan8 天前

Lower prices will help, but factory adoption will come down to reliability. A robot that works 95% of the time looks great in a demo and gets expensive fast during a real shift.

Dr. Paul De Santis, PharmD 的头像
Dr. Paul De Santis, PharmD9 天前

@Grok how much additional HBM memory will this require, present data for both the old and new estimates side by side

Zero G Talent 的头像
Zero G Talent7 天前

At that kind of projected deployment scale for humanoids, where is the engineering bottleneck going to hit first, the actuators or the power systems?

Kieu Van My Nghia 的头像
Kieu Van My Nghia7 天前

A rigorous analysis of cloud computing should consider scalability, reliability, and long-term sustainability over the next decade.

Milk Road AI 的头像
Milk Road AI9 天前

Harmonic Drive 👀

Melvin 的头像
Melvin9 天前

that's a good one but look at PE lol

FKA BIOS4L 的头像
FKA BIOS4L9 天前

$TSLA to the moon

Melvin 的头像
Melvin9 天前

I like Tesla but they need to start showing some results!

FKA BIOS4L 的头像
FKA BIOS4L9 天前

Agree. They will get it done but Elon has his hands full. Who do you see as the other players

Melvin 的头像
Melvin9 天前

too early to tell and majority of them are private. I am more so focused on the supply chain!

FKA BIOS4L 的头像
FKA BIOS4L9 天前

Thanks.

Julian Ashford 的头像
Julian Ashford9 天前

This analysis has real texture. Your perspective alongside @Wernerschnitzl’s research makes the broader market story more compelling.

Rachel Collins 的头像
Rachel Collins9 天前

Exactly the kind of analysis I come to X for. You and @Devorahsounds are both worth following.

相关视频

Robotics is following AI’s exact playbook and the Capex explosion is coming next. The first step is already happening, money is flooding into the companies building physical AI. Robotics and physical AI startups raised about $16.3 billion across 492 deals in the first quarter of 2026, roughly 4.5 times the average quarterly funding from 2021–2025. In the first half of 2026, physical AI companies raised $47.4 billion, more than the sector raised across all of 2022–2024 combined. That is how the AI cycle started, venture capital funded the technology first, then companies began spending hundreds of billions on the infrastructure needed to deploy it. Robotics is now moving from research labs into warehouses, factories, auto plants, logistics centers, and defense systems. Goldman Sachs raised its 2035 humanoid robot forecast from 1.38 million units to 6.48 million, with the market potentially reaching $138 billion. The reason is falling costs and Goldman expects average robot prices to decline from about $41,800 in 2025 to $21,300 by 2035. As prices fall, the payback period could shrink from 2.8 years in 2026 to about 1.9 years in 2027, making robots much easier for companies to justify as capital investments. That is when robotics can trigger its own capex cycle. Companies will spend not just on robots but also on factories, sensors, chips, batteries, software, power systems, data centers, and new automated facilities. Amazon is already a major example because its automation program could save roughly $72 billion between 2026 and 2030 and add about 240 basis points to operating margins. Morgan Stanley sees the long term opportunity as even larger, estimating 1 billion humanoid robots and about $7.5 trillion in annual revenue by 2050. And the biggest beneficiaries in all of this will be the picks and shovels companies behind the robots. That includes Nvidia and Renesas for chips, Teradyne for automation, Harmonic Drive for precision gearboxes, and Toyota, Honda, JTEKT, Aisin, and MinebeaMitsumi for manufacturing and motion control components. Bullish on robotics and the picks and shovels behind the next capex cycle and If you enjoyed reading this, make sure to follow Melvin for more robotics and AI insights. If you want to see exactly what I'm buying as an analyst at Milk Road Pro, you can join for just $1 using the link below.
3:02

Robotics is following AI’s exact playbook and the Capex explosion is coming next. The first step is already happening, money is flooding into the companies building physical AI. Robotics and physical AI startups raised about $16.3 billion across 492 deals in the first quarter of 2026, roughly 4.5 times the average quarterly funding from 2021–2025. In the first half of 2026, physical AI companies raised $47.4 billion, more than the sector raised across all of 2022–2024 combined. That is how the AI cycle started, venture capital funded the technology first, then companies began spending hundreds of billions on the infrastructure needed to deploy it. Robotics is now moving from research labs into warehouses, factories, auto plants, logistics centers, and defense systems. Goldman Sachs raised its 2035 humanoid robot forecast from 1.38 million units to 6.48 million, with the market potentially reaching $138 billion. The reason is falling costs and Goldman expects average robot prices to decline from about $41,800 in 2025 to $21,300 by 2035. As prices fall, the payback period could shrink from 2.8 years in 2026 to about 1.9 years in 2027, making robots much easier for companies to justify as capital investments. That is when robotics can trigger its own capex cycle. Companies will spend not just on robots but also on factories, sensors, chips, batteries, software, power systems, data centers, and new automated facilities. Amazon is already a major example because its automation program could save roughly $72 billion between 2026 and 2030 and add about 240 basis points to operating margins. Morgan Stanley sees the long term opportunity as even larger, estimating 1 billion humanoid robots and about $7.5 trillion in annual revenue by 2050. And the biggest beneficiaries in all of this will be the picks and shovels companies behind the robots. That includes Nvidia and Renesas for chips, Teradyne for automation, Harmonic Drive for precision gearboxes, and Toyota, Honda, JTEKT, Aisin, and MinebeaMitsumi for manufacturing and motion control components. Bullish on robotics and the picks and shovels behind the next capex cycle and If you enjoyed reading this, make sure to follow Melvin for more robotics and AI insights. If you want to see exactly what I'm buying as an analyst at Milk Road Pro, you can join for just $1 using the link below.

Melvin

23,055 次观看 • 8 天前

This is WILD! Morgan Stanley projects a $7.5 trillion humanoid robot market by 2050 roughly one robot for every six people alive on Earth today (Save this). That number is so large it is almost impossible to process so here is the context, the entire humanoid robot industry generated literally zero commercial revenue in 2024. In 2025 it generated approximately $3 billion and that gap between where this industry is today and where it is the investment opportunity. Every transformational technology industry looks embryonic at exactly the moment before it inflects, the internet in 1995, smartphones in 2007, cloud in 2010. In every case, the investors who looked at the current revenue and said this is too small to matter missed the entire compounding run. The investors who looked at the structural driver and asked what happens to every industry when this technology becomes cheap and ubiquitous, captured generational returns. The structural driver here is the global labor market, which is worth approximately $30 trillion annually. Morgan Stanley estimates that roughly 75% of occupations and 40% of all employees in the United States have some degree of humanoidability meaning their work could, at some price point, be performed by a humanoid robot. That addressable market in the US alone is approximately $3 trillion and globally, the prize is the largest labor replacement opportunity in the history of capitalism. The cost curve is the key that unlocks it all. A humanoid robot cost roughly $200,000 in 2024, Morgan Stanley projects that falls to approximately $150,000 by 2028, and to $50,000 by 2050 in high-income countries and as low as $15,000 where Chinese supply chains dominate. At $50,000, the economics become compelling across an enormous range of manufacturing, logistics, warehousing and service jobs. At $15,000, they become compelling for consumer applications as well and the cost trajectory is essentially the same one the semiconductor industry has followed for 60 years and it produces the same outcome every time. The hardware component market alone is expected to reach $780 billion by 2040 as a result. The demand multipliers for specific components by 2050 compared to 2025 levels are staggering, edge compute chips up 1,904x, bearings up 370x, motors up 170x and precision reducers up 157x. Every one of those supply chain layers represents a distinct investment category that barely exists in meaningful scale today. Bullish on Robotics and make sure to follow me Melvin for more overlooked opportunities in robotics.
4:22

This is WILD! Morgan Stanley projects a $7.5 trillion humanoid robot market by 2050 roughly one robot for every six people alive on Earth today (Save this). That number is so large it is almost impossible to process so here is the context, the entire humanoid robot industry generated literally zero commercial revenue in 2024. In 2025 it generated approximately $3 billion and that gap between where this industry is today and where it is the investment opportunity. Every transformational technology industry looks embryonic at exactly the moment before it inflects, the internet in 1995, smartphones in 2007, cloud in 2010. In every case, the investors who looked at the current revenue and said this is too small to matter missed the entire compounding run. The investors who looked at the structural driver and asked what happens to every industry when this technology becomes cheap and ubiquitous, captured generational returns. The structural driver here is the global labor market, which is worth approximately $30 trillion annually. Morgan Stanley estimates that roughly 75% of occupations and 40% of all employees in the United States have some degree of humanoidability meaning their work could, at some price point, be performed by a humanoid robot. That addressable market in the US alone is approximately $3 trillion and globally, the prize is the largest labor replacement opportunity in the history of capitalism. The cost curve is the key that unlocks it all. A humanoid robot cost roughly $200,000 in 2024, Morgan Stanley projects that falls to approximately $150,000 by 2028, and to $50,000 by 2050 in high-income countries and as low as $15,000 where Chinese supply chains dominate. At $50,000, the economics become compelling across an enormous range of manufacturing, logistics, warehousing and service jobs. At $15,000, they become compelling for consumer applications as well and the cost trajectory is essentially the same one the semiconductor industry has followed for 60 years and it produces the same outcome every time. The hardware component market alone is expected to reach $780 billion by 2040 as a result. The demand multipliers for specific components by 2050 compared to 2025 levels are staggering, edge compute chips up 1,904x, bearings up 370x, motors up 170x and precision reducers up 157x. Every one of those supply chain layers represents a distinct investment category that barely exists in meaningful scale today. Bullish on Robotics and make sure to follow me Melvin for more overlooked opportunities in robotics.

Melvin

15,545 次观看 • 2 个月前

Micron is going to $4,000 and this exactly why (Save this). Hyperscaler Capex, the combined spending of Amazon, Google, Meta, Microsoft and Oracle was $261 billion in 2024 and it hit $449 billion in 2025. Morgan Stanley now expects $805 billion in 2026 and $1.1 trillion in 2027. Memory consistently runs at 35–48% of that total hardware spend, apply that range to Morgan Stanley's numbers and you get somewhere between $280 billion and $530 billion flowing into memory stocks over the next two years alone. That is the market Micron is selling into right now and the company just reported $41.5 billion in revenue in a single quarter with 85% gross margins. But data centers are only the first wave. L2+ vehicles, cars with meaningful driver assistance carry over five times the memory of a standard car and that mix is doubling to over 20% of all vehicles sold this year, and Micron expects it to hit 40% by 2030. Autonomous vehicles will require over 300 gigabytes of DRAM per car, an 18x increase in memory content per unit, applied across tens of millions of cars a year. The third wave is the one that makes automotive look small, humanoid robots. A humanoid robot carries 10 times the memory of an average L2 vehicle. Tesla, Figure, and a growing list of US robotics companies are still in the very early stages while China is already scaling humanoid production fast. When the US robotics boom arrives and it will, the memory requirement per unit is orders of magnitude larger than anything that has come before. The one risk worth naming is if the hyperscalers signal a pause in spending in 2027, that puts real pressure on the thesis. But Morgan Stanley has raised their capex forecast by $630 billion in six months alone. The data center boom is already here,the car boom is arriving and the robotics boom hasn't started yet. Micron is the only US-based company that can supply all three. Follow me Melvin for more AI, semis and the next big market themes.
3:25

Micron is going to $4,000 and this exactly why (Save this). Hyperscaler Capex, the combined spending of Amazon, Google, Meta, Microsoft and Oracle was $261 billion in 2024 and it hit $449 billion in 2025. Morgan Stanley now expects $805 billion in 2026 and $1.1 trillion in 2027. Memory consistently runs at 35–48% of that total hardware spend, apply that range to Morgan Stanley's numbers and you get somewhere between $280 billion and $530 billion flowing into memory stocks over the next two years alone. That is the market Micron is selling into right now and the company just reported $41.5 billion in revenue in a single quarter with 85% gross margins. But data centers are only the first wave. L2+ vehicles, cars with meaningful driver assistance carry over five times the memory of a standard car and that mix is doubling to over 20% of all vehicles sold this year, and Micron expects it to hit 40% by 2030. Autonomous vehicles will require over 300 gigabytes of DRAM per car, an 18x increase in memory content per unit, applied across tens of millions of cars a year. The third wave is the one that makes automotive look small, humanoid robots. A humanoid robot carries 10 times the memory of an average L2 vehicle. Tesla, Figure, and a growing list of US robotics companies are still in the very early stages while China is already scaling humanoid production fast. When the US robotics boom arrives and it will, the memory requirement per unit is orders of magnitude larger than anything that has come before. The one risk worth naming is if the hyperscalers signal a pause in spending in 2027, that puts real pressure on the thesis. But Morgan Stanley has raised their capex forecast by $630 billion in six months alone. The data center boom is already here,the car boom is arriving and the robotics boom hasn't started yet. Micron is the only US-based company that can supply all three. Follow me Melvin for more AI, semis and the next big market themes.

Melvin

27,296 次观看 • 2 个月前

China's humanoid robotics market is on fire. With orders expected to top 30,000 units this year—a tenfold jump from 2024's total of less than 3,000—2025 is officially shaping up to be the "Year of Mass Production." This surge, driven by an expansion into new sectors like industrial manufacturing, logistics, and elder care, is reflected in a wave of new deals across the industry. Here's a look at some of the key commercial progress: Astribot: A 1,000-unit order for industrial and logistics deployment over two years. TianTai Robotics: Signed a major 10,000-unit order for caregiving robots. Noetix Robotics : Received over 2,000 intent orders in one month, valued at over 100 million yuan, with a focus on education and commercial performances. AgiBot: Expects to ship thousands of units this year and tens of thousands in 2026. Unitree Robotics: Has orders for thousands of units and is one of the most visible products in the industry. UBTech: Aims to deliver 500 industrial humanoids in 2025, with educational robot orders already exceeding 300 units. Robot Era: Delivered over 300 units by July 2025 with 500 more on hand. TLIBOT: Has around 1,000 intent orders. Galbot: Secured orders for its supermarket security robot, Galbot, in 100 stores. AI² Robotics: Has nearly 500 orders for its general-purpose robots for industrial and public service scenarios. But here’s the crucial reality check. While the order boom is exciting, it doesn't automatically translate to fulfilled deliveries. Many companies lack the production capacity to keep up. A significant portion of these are "intent orders" or framework agreements, not guaranteed sales. Furthermore, the market is heavily B2B-focused, with consumer demand representing only about 5% of sales. Some orders are even symbolic, for public relations or strategic purposes. This “order frenzy” is a starting point, not the finish line. The true test for China's humanoid robot industry isn't who can secure the biggest order, but who can consistently deliver on it and build a stable market for the future.
2:44

China's humanoid robotics market is on fire. With orders expected to top 30,000 units this year—a tenfold jump from 2024's total of less than 3,000—2025 is officially shaping up to be the "Year of Mass Production." This surge, driven by an expansion into new sectors like industrial manufacturing, logistics, and elder care, is reflected in a wave of new deals across the industry. Here's a look at some of the key commercial progress: Astribot: A 1,000-unit order for industrial and logistics deployment over two years. TianTai Robotics: Signed a major 10,000-unit order for caregiving robots. Noetix Robotics : Received over 2,000 intent orders in one month, valued at over 100 million yuan, with a focus on education and commercial performances. AgiBot: Expects to ship thousands of units this year and tens of thousands in 2026. Unitree Robotics: Has orders for thousands of units and is one of the most visible products in the industry. UBTech: Aims to deliver 500 industrial humanoids in 2025, with educational robot orders already exceeding 300 units. Robot Era: Delivered over 300 units by July 2025 with 500 more on hand. TLIBOT: Has around 1,000 intent orders. Galbot: Secured orders for its supermarket security robot, Galbot, in 100 stores. AI² Robotics: Has nearly 500 orders for its general-purpose robots for industrial and public service scenarios. But here’s the crucial reality check. While the order boom is exciting, it doesn't automatically translate to fulfilled deliveries. Many companies lack the production capacity to keep up. A significant portion of these are "intent orders" or framework agreements, not guaranteed sales. Furthermore, the market is heavily B2B-focused, with consumer demand representing only about 5% of sales. Some orders are even symbolic, for public relations or strategic purposes. This “order frenzy” is a starting point, not the finish line. The true test for China's humanoid robot industry isn't who can secure the biggest order, but who can consistently deliver on it and build a stable market for the future.

RoboHub🤖

199,146 次观看 • 1 年前

Nvidia's next generation chips are about to make a handful of companies impossible to ignore and here is how you can benefit from this (Save this). Goldman projects networking content per AI compute system could rise from $315,000 today to $9.4 million in Nvidia's next generation, a 29x increase in what companies spend just wiring their chips together. Morgan Stanley, Citi and Goldman all agree on the same direction, projecting the total addressable market for AI networking to grow from around $11 to $15 billion today to $154 billion by 2028, roughly a 9x jump in just a few years. As AI clusters scale into the hundreds of thousands of GPUs, the chips themselves stop being the bottleneck and the connections between them become the limiting factor instead. Every GPU needs to talk to every other GPU almost instantly to keep a training run synchronized and that requires far more sophisticated switches, optical modules and cabling than today's networks can provide, which is exactly why Nvidia's next generation systems are expected to need dramatically more networking hardware per system just to keep up. Optical modules specifically are becoming the chokepoint. The global optical module total addressable market is projected to grow from roughly $6.7 billion in the first quarter of 2025 to nearly $69 billion by 2028, with 1.6T modules going from essentially zero market share today to capturing the majority of shipments by 2028 as speeds keep climbing. Nvidia sits at the center of this story since it designs the switches and networking silicon that go into every one of these systems, meaning that 29x jump in networking content per system flows straight into Nvidia's own revenue per AI factory, not just its GPU sales. Marvell is one of the biggest direct beneficiaries here, since it makes the custom networking chips and interconnect silicon that hyperscalers rely on to move data between GPUs at these speeds. Credo Technology plays a slightly different role, supplying the high speed connectivity chips that clean up signals moving through copper links, which becomes more valuable as networks push toward higher bandwidth with lower power draw. AAOI, or Applied Optoelectronics, sits closer to the physical optics layer, making the lasers and optical components that go into the transceivers carrying data across these networks, positioning it directly in the path of that jump from $6.7 billion to $69 billion in optical module demand. Coherent and Lumentum round out the optical side, both supplying lasers and photonic components that scale directly with 1.6T module adoption as it goes from near zero to the majority of shipments by 2028. Switch and connector makers benefit almost mechanically from this trend too, since every additional dollar of networking content per system has to physically pass through a switch, a cable, or a connector, meaning revenue scales with network complexity regardless of which chipmaker's GPUs sit inside the rack. Bullish on AI networking, make sure to follow Melvin for more semiconductor insights and check out the link below for more details.
3:38

Nvidia's next generation chips are about to make a handful of companies impossible to ignore and here is how you can benefit from this (Save this). Goldman projects networking content per AI compute system could rise from $315,000 today to $9.4 million in Nvidia's next generation, a 29x increase in what companies spend just wiring their chips together. Morgan Stanley, Citi and Goldman all agree on the same direction, projecting the total addressable market for AI networking to grow from around $11 to $15 billion today to $154 billion by 2028, roughly a 9x jump in just a few years. As AI clusters scale into the hundreds of thousands of GPUs, the chips themselves stop being the bottleneck and the connections between them become the limiting factor instead. Every GPU needs to talk to every other GPU almost instantly to keep a training run synchronized and that requires far more sophisticated switches, optical modules and cabling than today's networks can provide, which is exactly why Nvidia's next generation systems are expected to need dramatically more networking hardware per system just to keep up. Optical modules specifically are becoming the chokepoint. The global optical module total addressable market is projected to grow from roughly $6.7 billion in the first quarter of 2025 to nearly $69 billion by 2028, with 1.6T modules going from essentially zero market share today to capturing the majority of shipments by 2028 as speeds keep climbing. Nvidia sits at the center of this story since it designs the switches and networking silicon that go into every one of these systems, meaning that 29x jump in networking content per system flows straight into Nvidia's own revenue per AI factory, not just its GPU sales. Marvell is one of the biggest direct beneficiaries here, since it makes the custom networking chips and interconnect silicon that hyperscalers rely on to move data between GPUs at these speeds. Credo Technology plays a slightly different role, supplying the high speed connectivity chips that clean up signals moving through copper links, which becomes more valuable as networks push toward higher bandwidth with lower power draw. AAOI, or Applied Optoelectronics, sits closer to the physical optics layer, making the lasers and optical components that go into the transceivers carrying data across these networks, positioning it directly in the path of that jump from $6.7 billion to $69 billion in optical module demand. Coherent and Lumentum round out the optical side, both supplying lasers and photonic components that scale directly with 1.6T module adoption as it goes from near zero to the majority of shipments by 2028. Switch and connector makers benefit almost mechanically from this trend too, since every additional dollar of networking content per system has to physically pass through a switch, a cable, or a connector, meaning revenue scales with network complexity regardless of which chipmaker's GPUs sit inside the rack. Bullish on AI networking, make sure to follow Melvin for more semiconductor insights and check out the link below for more details.

Melvin

42,220 次观看 • 1 个月前

Nebius will be a trillion dollar company (Save this). The neocloud market, purpose-built AI cloud infrastructure, separate from legacy hyperscalers generated roughly $25 billion in revenue in 2025, up 223% year over year. Synergy Research projects it will approach $400 billion by 2031, compounding at 58% annually one of the fastest sustained growth rates ever recorded for an infrastructure category of this scale. The CEO's explanation for why they win is worth understanding in detail. GPU compute is scarce and that part everyone knows but Nebius is not simply renting GPUs by the hour and marking them up, which is what most neocloud imitators do. They have built their own physical capacity for inference, optimized the full technology stack from the software layer all the way down to the rack hardware and recently acquired a company called Agen specifically to push inference latency even lower and throughput even higher. The CEO frames the core problem directly that in 2026, every product you build is powered by tokens, AI intelligence and while you can get those tokens from OpenAI or Anthropic via a simple API call, the moment you want to run open source models, specialized vertical models, or anything other than the two dominant frontier labs, you run into a wall. You can download the weights from Hugging Face and assemble the pieces. But getting those workloads to run at scale, at the economics you need, with the reliability your product requires, is an extraordinarily complex engineering challenge that most companies cannot staff or afford to solve in-house. That is the problem Nebius is solving, and that is why their inference product called Token Factory exists. The financial results are among the most dramatic growth numbers reported by any public company this year. In Q1 2026, Nebius posted $399 million in revenue, a 684% increase from the same quarter a year earlier. In the span of twelve months, the company swung from a $104 million net loss to $621 million in net income. Cash from operations went from negative $184 million to positive $2.26 billion in the same period meaning this is not growth funded by burning investor capital, it is growth that is now generating its own fuel. For the full year 2026, Nebius is guiding for an annualized revenue run rate of $7 billion to $9 billion, with pipeline creation tracking to surpass $4 billion. The contracted backlog sits at $49 billion, anchored by a $27 billion agreement with Meta, a deal worth up to $19.4 billion with Microsoft, and a public endorsement from Jensen Huang at NVIDIA's GTC conference in 2026. The current market cap is approximately $56 billion. A company with $7 to $9 billion in annualized revenue, growing at 684%, turning cash-flow positive, sitting on $49 billion in contracted backlog, operating in a market compounding at 58% annually toward $400 billion, that company has a credible path to 20x from its current valuation if execution holds. That is the trillion dollar case, and it does not require any heroic assumptions and it requires Nebius to keep doing what it is already demonstrably doing. Milk Road Pro called this one early. Our analysts added Nebius to the portfolio when it was still flying under the radar, and we are sitting on a massive gain on that position right now. If you want to see what else we are building conviction on before the rest of the market catches up, come join us at Milk Road Pro using the link below!
1:04

Nebius will be a trillion dollar company (Save this). The neocloud market, purpose-built AI cloud infrastructure, separate from legacy hyperscalers generated roughly $25 billion in revenue in 2025, up 223% year over year. Synergy Research projects it will approach $400 billion by 2031, compounding at 58% annually one of the fastest sustained growth rates ever recorded for an infrastructure category of this scale. The CEO's explanation for why they win is worth understanding in detail. GPU compute is scarce and that part everyone knows but Nebius is not simply renting GPUs by the hour and marking them up, which is what most neocloud imitators do. They have built their own physical capacity for inference, optimized the full technology stack from the software layer all the way down to the rack hardware and recently acquired a company called Agen specifically to push inference latency even lower and throughput even higher. The CEO frames the core problem directly that in 2026, every product you build is powered by tokens, AI intelligence and while you can get those tokens from OpenAI or Anthropic via a simple API call, the moment you want to run open source models, specialized vertical models, or anything other than the two dominant frontier labs, you run into a wall. You can download the weights from Hugging Face and assemble the pieces. But getting those workloads to run at scale, at the economics you need, with the reliability your product requires, is an extraordinarily complex engineering challenge that most companies cannot staff or afford to solve in-house. That is the problem Nebius is solving, and that is why their inference product called Token Factory exists. The financial results are among the most dramatic growth numbers reported by any public company this year. In Q1 2026, Nebius posted $399 million in revenue, a 684% increase from the same quarter a year earlier. In the span of twelve months, the company swung from a $104 million net loss to $621 million in net income. Cash from operations went from negative $184 million to positive $2.26 billion in the same period meaning this is not growth funded by burning investor capital, it is growth that is now generating its own fuel. For the full year 2026, Nebius is guiding for an annualized revenue run rate of $7 billion to $9 billion, with pipeline creation tracking to surpass $4 billion. The contracted backlog sits at $49 billion, anchored by a $27 billion agreement with Meta, a deal worth up to $19.4 billion with Microsoft, and a public endorsement from Jensen Huang at NVIDIA's GTC conference in 2026. The current market cap is approximately $56 billion. A company with $7 to $9 billion in annualized revenue, growing at 684%, turning cash-flow positive, sitting on $49 billion in contracted backlog, operating in a market compounding at 58% annually toward $400 billion, that company has a credible path to 20x from its current valuation if execution holds. That is the trillion dollar case, and it does not require any heroic assumptions and it requires Nebius to keep doing what it is already demonstrably doing. Milk Road Pro called this one early. Our analysts added Nebius to the portfolio when it was still flying under the radar, and we are sitting on a massive gain on that position right now. If you want to see what else we are building conviction on before the rest of the market catches up, come join us at Milk Road Pro using the link below!

Milk Road AI

28,622 次观看 • 3 个月前

Morgan Stanley just raised their 2027 AI capex forecast to $1.1 trillion and that number still doesn't include SpaceX or a lot of the other AI companies (Save this). When you factor those in, the real 2027 figure is probably closer to $1.5 trillion and AI lab inference revenue combined is tracking toward $300 billion in 2027. On its surface that ratio sounds alarming, spending $1.5 trillion in capex to generate $300 billion in revenue. But the framing collapses the moment you examine two things the bears consistently ignore, gross margins and the revenue trajectory. Gross margins on inference revenue are running at 60 to 70 percent. That means the $300 billion in inference revenue generates $180 to $210 billion in gross profit and that number compounds rapidly as utilization scales on infrastructure that is already built and paid for. The Capex is not being deployed against today's revenue but rather being deployed against a revenue trajectory that has shown no signs of decelerating. To understand how aggressive that trajectory actually is, consider that Morgan Stanley's $1.1 trillion hyperscaler forecast is nearly double what analysts projected for the same year just twelve months ago And they described the demand as inelastic, meaning it is not slowing down regardless of rising costs, tighter financing conditions or geopolitical risk. The AI industry ended 2025 tracking well over $200 billion in combined inference revenue and the growth rate since then has continued to accelerate rather than flatten. Anthropic alone scaled from negligible revenue to a $30 billion annualized run rate in approximately 18 months while OpenAI is tracking toward $280 billion in annual revenue by 2030 from $13 billion in 2025. There is also a structural reality in the capex number that the bears never account for. Roughly 35 percent of total AI spending goes toward training, building the next model generation which is not revenue-generating in the current period. That means only about 65 percent of the $1.5 trillion in capex is actually deployed against the inference infrastructure that earns revenue today. When you apply the 60 to 70 percent gross margin to the revenue that sits on top of that 65 percent figure, the economics look substantially better than the headline capex to revenue ratio implies. Every CEO who has been closest to this buildout has consistently underestimated it and Jensen Huang projected $1 trillion in AI capex two years ago and was called delusional. Dario Amodei said in early 2026 that AI revenues would reach the low hundreds of billions by 2028 and trillions before 2030 and given where Anthropic's own revenue trajectory is today, he is likely revising those numbers upward. The pattern here is consistent, every time someone models the revenue ceiling, the actual number breaks through it faster than expected. Come join Milk Road Pro for our full breakdown, the real unit economics of the AI inference buildout, how the capex to revenue ratio evolves over the next three years, and our entire AI thesis! Link below!
4:04

Morgan Stanley just raised their 2027 AI capex forecast to $1.1 trillion and that number still doesn't include SpaceX or a lot of the other AI companies (Save this). When you factor those in, the real 2027 figure is probably closer to $1.5 trillion and AI lab inference revenue combined is tracking toward $300 billion in 2027. On its surface that ratio sounds alarming, spending $1.5 trillion in capex to generate $300 billion in revenue. But the framing collapses the moment you examine two things the bears consistently ignore, gross margins and the revenue trajectory. Gross margins on inference revenue are running at 60 to 70 percent. That means the $300 billion in inference revenue generates $180 to $210 billion in gross profit and that number compounds rapidly as utilization scales on infrastructure that is already built and paid for. The Capex is not being deployed against today's revenue but rather being deployed against a revenue trajectory that has shown no signs of decelerating. To understand how aggressive that trajectory actually is, consider that Morgan Stanley's $1.1 trillion hyperscaler forecast is nearly double what analysts projected for the same year just twelve months ago And they described the demand as inelastic, meaning it is not slowing down regardless of rising costs, tighter financing conditions or geopolitical risk. The AI industry ended 2025 tracking well over $200 billion in combined inference revenue and the growth rate since then has continued to accelerate rather than flatten. Anthropic alone scaled from negligible revenue to a $30 billion annualized run rate in approximately 18 months while OpenAI is tracking toward $280 billion in annual revenue by 2030 from $13 billion in 2025. There is also a structural reality in the capex number that the bears never account for. Roughly 35 percent of total AI spending goes toward training, building the next model generation which is not revenue-generating in the current period. That means only about 65 percent of the $1.5 trillion in capex is actually deployed against the inference infrastructure that earns revenue today. When you apply the 60 to 70 percent gross margin to the revenue that sits on top of that 65 percent figure, the economics look substantially better than the headline capex to revenue ratio implies. Every CEO who has been closest to this buildout has consistently underestimated it and Jensen Huang projected $1 trillion in AI capex two years ago and was called delusional. Dario Amodei said in early 2026 that AI revenues would reach the low hundreds of billions by 2028 and trillions before 2030 and given where Anthropic's own revenue trajectory is today, he is likely revising those numbers upward. The pattern here is consistent, every time someone models the revenue ceiling, the actual number breaks through it faster than expected. Come join Milk Road Pro for our full breakdown, the real unit economics of the AI inference buildout, how the capex to revenue ratio evolves over the next three years, and our entire AI thesis! Link below!

Milk Road AI

21,141 次观看 • 3 个月前

Everything Elon said about Optimus at the All-In Summit today: • We’re finalizing the design of Optimus v3. That release is going to be a very remarkable robot. It will have manual dexterity comparable to a human, meaning a very complex hand, an AI mind that can navigate and comprehend reality, and will be made in very high volume. • Other robotics companies are missing those three very hard things. • I spend more mental cycles on Optimus than any other single thing. Solving real-world AI, all of the electrical-mechanical issues, the supply chain, and production challenges. • There is no supply chain for humanoid robots, so it has to be created from scratch, which requires a lot of vertical integration. None of the actuators in Optimus are available from an existing supply chain. • I think if successful, Optimus would be the biggest product ever. • The marginal cost of production, once we hit a million units per year, will probably be around $20,000. It depends on how much we spend on the AI chip in the robot, and we’ll need to achieve a lot of efficiencies in the actuators—26 actuators per arm (26 motors, gearboxes, and power electronics). The AI chip might cost $5,000 or $6,000, maybe more. At 1 million units a year, production cost will be $20,000, maybe $25,000. Price will be a function of demand. • Human hands have evolved to be incredibly sophisticated machines. Hands are a very first instrument. You can swing a baseball bat, thread a needle, play a piano or violin, and assemble a car. Hands are incredibly versatile instruments. Most of the muscles of the hands are actually in the forearm, and the hand is almost like a puppet. Human tendon evolution is incredibly good. The human hand has 27 or 28 degrees of freedom, depending on how you count it; it’s amazing. • In order to create a robot that can be a generalized humanoid, you must solve the “hands problem.” • Even though there are 10,000 to 20,000 electric motors out there, we couldn’t buy the actuators for any amount of money. We had to design every electric motor, gearbox, and controlling electronics from scratch, from first principles of physics. • Optimus is harder than developing any previous Tesla product, but not harder than Starship. • Right now, we’re struggling with the final design of the hardware, primarily the hand. The hands and forearm are the majority of the engineering difficulty of the entire robot. • If you want to do all the things that a human can do, it turns out you need a humanoid robot. If you want to do a subset, that’s much easier. Humans evolved to the shape and capability that we have for a good reason. There is value to having four fingers and a thumb; even the pinky is quite useful. Toes are much more of a question mark. • The AI5 inference chip will be 40 times better than AI4 by some measures. We know the limiting factors of the chip because the AI software and hardware teams work so closely. Effectively, the Tesla AI hardware and software teams are co-designing the chip. • The Softmax function on AI4 takes 40 steps in emulation mode, which will take only a few steps in AI5 natively. AI5 will easily handle mixed precision. • In terms of nominal raw compute, the AI5 inference chip has 8 times more compute, 9 times more memory, and 5 times more memory bandwidth compared to AI4. Because we’re addressing some core limitations and optimizations at the silicon level, we’re able to realize 40x improvements.
12:26

Everything Elon said about Optimus at the All-In Summit today: • We’re finalizing the design of Optimus v3. That release is going to be a very remarkable robot. It will have manual dexterity comparable to a human, meaning a very complex hand, an AI mind that can navigate and comprehend reality, and will be made in very high volume. • Other robotics companies are missing those three very hard things. • I spend more mental cycles on Optimus than any other single thing. Solving real-world AI, all of the electrical-mechanical issues, the supply chain, and production challenges. • There is no supply chain for humanoid robots, so it has to be created from scratch, which requires a lot of vertical integration. None of the actuators in Optimus are available from an existing supply chain. • I think if successful, Optimus would be the biggest product ever. • The marginal cost of production, once we hit a million units per year, will probably be around $20,000. It depends on how much we spend on the AI chip in the robot, and we’ll need to achieve a lot of efficiencies in the actuators—26 actuators per arm (26 motors, gearboxes, and power electronics). The AI chip might cost $5,000 or $6,000, maybe more. At 1 million units a year, production cost will be $20,000, maybe $25,000. Price will be a function of demand. • Human hands have evolved to be incredibly sophisticated machines. Hands are a very first instrument. You can swing a baseball bat, thread a needle, play a piano or violin, and assemble a car. Hands are incredibly versatile instruments. Most of the muscles of the hands are actually in the forearm, and the hand is almost like a puppet. Human tendon evolution is incredibly good. The human hand has 27 or 28 degrees of freedom, depending on how you count it; it’s amazing. • In order to create a robot that can be a generalized humanoid, you must solve the “hands problem.” • Even though there are 10,000 to 20,000 electric motors out there, we couldn’t buy the actuators for any amount of money. We had to design every electric motor, gearbox, and controlling electronics from scratch, from first principles of physics. • Optimus is harder than developing any previous Tesla product, but not harder than Starship. • Right now, we’re struggling with the final design of the hardware, primarily the hand. The hands and forearm are the majority of the engineering difficulty of the entire robot. • If you want to do all the things that a human can do, it turns out you need a humanoid robot. If you want to do a subset, that’s much easier. Humans evolved to the shape and capability that we have for a good reason. There is value to having four fingers and a thumb; even the pinky is quite useful. Toes are much more of a question mark. • The AI5 inference chip will be 40 times better than AI4 by some measures. We know the limiting factors of the chip because the AI software and hardware teams work so closely. Effectively, the Tesla AI hardware and software teams are co-designing the chip. • The Softmax function on AI4 takes 40 steps in emulation mode, which will take only a few steps in AI5 natively. AI5 will easily handle mixed precision. • In terms of nominal raw compute, the AI5 inference chip has 8 times more compute, 9 times more memory, and 5 times more memory bandwidth compared to AI4. Because we’re addressing some core limitations and optimizations at the silicon level, we’re able to realize 40x improvements.

The Humanoid Hub

239,049 次观看 • 1 年前

Chamath has been watching SpaceX for 15 years and he thinks the market is still not close to understanding what it actually is (Save this). The first argument is the industrial logic of a Tesla SpaceX combination. One capital structure, one balance sheet, one vehicle to raise money across robotics, autonomous vehicles, energy, AI, and launch. Chamath Palihapitiya argument is that markets are treating this as a peripheral possibility rather than an obvious strategic inevitability. The second is Starlink Direct to Cell, which he believes will generate enormous domestic cellular revenue before most of the bigger SpaceX narratives even begin to materialize. The numbers already back this up. Starlink has over 10 million Direct to Cell monthly active users with live partnerships with T-Mobile, Rogers and Optus standard smartphones connecting directly to satellites with no special hardware required. SpaceX is currently deploying approximately 340 Direct to Cell satellites per month, targeting 25 million monthly active users by end of 2026. Goldman forecasts SpaceX's AI division will generate $15.6 billion in 2026, rising to $34.5 billion in 2027 and accelerating to $322 billion by 2030 roughly a 100-fold increase in five years. Total SpaceX revenue hits $474 billion by 2030, up from $18.7 billion in 2025. The launch cadence numbers are where this gets staggering. SpaceX is expected to execute 151 Starship launches in 2027, scaling to 253 in 2028, then 1,504 in 2029, 2,808 in 2030, and 5,467 in 2031. Goldman projects 5,288 of those 2031 launches will be dedicated Starship AI missions each carrying 30 to 50 satellites powered by one GB300 equivalent compute rack apiece. The cost per kilogram to orbit falls below $100 as reusability matures, compared to $1,500 per kilogram on Falcon 9 today. Morgan Stanley projected a 24-hour turnaround by late 2027, enabling the kind of cadence these numbers require. That launch cost collapse is what makes the orbital AI compute thesis real Elon Musk
1:08

Chamath has been watching SpaceX for 15 years and he thinks the market is still not close to understanding what it actually is (Save this). The first argument is the industrial logic of a Tesla SpaceX combination. One capital structure, one balance sheet, one vehicle to raise money across robotics, autonomous vehicles, energy, AI, and launch. Chamath Palihapitiya argument is that markets are treating this as a peripheral possibility rather than an obvious strategic inevitability. The second is Starlink Direct to Cell, which he believes will generate enormous domestic cellular revenue before most of the bigger SpaceX narratives even begin to materialize. The numbers already back this up. Starlink has over 10 million Direct to Cell monthly active users with live partnerships with T-Mobile, Rogers and Optus standard smartphones connecting directly to satellites with no special hardware required. SpaceX is currently deploying approximately 340 Direct to Cell satellites per month, targeting 25 million monthly active users by end of 2026. Goldman forecasts SpaceX's AI division will generate $15.6 billion in 2026, rising to $34.5 billion in 2027 and accelerating to $322 billion by 2030 roughly a 100-fold increase in five years. Total SpaceX revenue hits $474 billion by 2030, up from $18.7 billion in 2025. The launch cadence numbers are where this gets staggering. SpaceX is expected to execute 151 Starship launches in 2027, scaling to 253 in 2028, then 1,504 in 2029, 2,808 in 2030, and 5,467 in 2031. Goldman projects 5,288 of those 2031 launches will be dedicated Starship AI missions each carrying 30 to 50 satellites powered by one GB300 equivalent compute rack apiece. The cost per kilogram to orbit falls below $100 as reusability matures, compared to $1,500 per kilogram on Falcon 9 today. Morgan Stanley projected a 24-hour turnaround by late 2027, enabling the kind of cadence these numbers require. That launch cost collapse is what makes the orbital AI compute thesis real Elon Musk

Milk Road AI

96,048 次观看 • 2 个月前

In 2025, China’s automobile production and sales reached 34.5 million units and 34.4 million units respectively. What does this set of data really mean? Based on China’s current 556 million licensed drivers, on average, 1 out of every 16 drivers got a new vehicle last year. China is essentially a massive auto factory combined with an enormous auto market. Annual production and sales have fluctuated but stayed nearly in balance, with demand even outstripping supply in some years. Take the past five years as an example: total auto production exceeded total sales by only 8,000 units, resulting in an inventory rate of just 0.05‰. The Chinese auto market is far from saturated, although the data has hit a new all-time high and topped the global charts for 17 straight years. In China, there are just 250 vehicles per 1,000 people—less than one-third of the U.S. level and half that of Europe or Japan. Right now, Chinese automobiles—especially new energy vehicles—are rapidly spreading into rural areas across the country. Rural buyers often opt for affordable NEVs priced around $7,000 for daily commuting or to make extra money driving for ride-hailing services. There are also plenty of models tailored for agricultural use, like new energy cargo vans and light trucks.
1:44

In 2025, China’s automobile production and sales reached 34.5 million units and 34.4 million units respectively. What does this set of data really mean? Based on China’s current 556 million licensed drivers, on average, 1 out of every 16 drivers got a new vehicle last year. China is essentially a massive auto factory combined with an enormous auto market. Annual production and sales have fluctuated but stayed nearly in balance, with demand even outstripping supply in some years. Take the past five years as an example: total auto production exceeded total sales by only 8,000 units, resulting in an inventory rate of just 0.05‰. The Chinese auto market is far from saturated, although the data has hit a new all-time high and topped the global charts for 17 straight years. In China, there are just 250 vehicles per 1,000 people—less than one-third of the U.S. level and half that of Europe or Japan. Right now, Chinese automobiles—especially new energy vehicles—are rapidly spreading into rural areas across the country. Rural buyers often opt for affordable NEVs priced around $7,000 for daily commuting or to make extra money driving for ride-hailing services. There are also plenty of models tailored for agricultural use, like new energy cargo vans and light trucks.

Sinical

95,774 次观看 • 7 个月前

Unitree Robotics just filed to go public on March 20th targeting a $7 billion valuation. Most people have no idea what this company actually is. Here is why this might be the most important robotics IPO of the decade. Unitree shipped 5,500 humanoid robots in 2025. Figure AI shipped roughly 150. Agility Robotics shipped roughly 150. Unitree did $246 million in revenue last year, up 335% year over year, and they are actually profitable. Figure AI is valued at $39 billion with near zero revenue and is still private. Unitree wants $7 billion with real numbers. The price point is what separates them from everyone else. Their G1 humanoid sells for $13,500. Competitors charge $50,000 to $130,000. Their newest R1 humanoid launching in April starts at $4,900. Nothing comparable exists at that price anywhere in the world. They hold roughly 32% of the global humanoid market and 70% of the quadruped market. The moat is vertical integration. They self develop over 90% of core components including motors, reducers, controllers, sensors, and all software. Real clients include PetroChina, Sinopec, State Grid, and China Mobile. This is not a research project. Product is shipping at scale. This is listing on China's STAR Market, not Hong Kong, not the US, which makes access extremely difficult for international investors. The risks are real. The US House Select Committee on the CCP has formally requested Unitree be blacklisted. Their robots appeared in PLA military exercises in 2024. Tariffs have already nearly tripled the US price of the G1. $TSLA Optimus is targeting sub $20,000 pricing with automotive scale manufacturing backed by $NVDA compute. If they execute, the price advantage shrinks fast. But this is still the only profitable pure play humanoid robotics company in the world growing at 335% a year, valued at a fraction of its loss making peers. Goldman projects the humanoid market at $38 billion by 2035. Morgan Stanley goes to $5 trillion by 2050. Unitree currently holds the largest market share of any humanoid manufacturer on the planet. Full breakdown coming soon. $TSLA $NVDA
0:44

Unitree Robotics just filed to go public on March 20th targeting a $7 billion valuation. Most people have no idea what this company actually is. Here is why this might be the most important robotics IPO of the decade. Unitree shipped 5,500 humanoid robots in 2025. Figure AI shipped roughly 150. Agility Robotics shipped roughly 150. Unitree did $246 million in revenue last year, up 335% year over year, and they are actually profitable. Figure AI is valued at $39 billion with near zero revenue and is still private. Unitree wants $7 billion with real numbers. The price point is what separates them from everyone else. Their G1 humanoid sells for $13,500. Competitors charge $50,000 to $130,000. Their newest R1 humanoid launching in April starts at $4,900. Nothing comparable exists at that price anywhere in the world. They hold roughly 32% of the global humanoid market and 70% of the quadruped market. The moat is vertical integration. They self develop over 90% of core components including motors, reducers, controllers, sensors, and all software. Real clients include PetroChina, Sinopec, State Grid, and China Mobile. This is not a research project. Product is shipping at scale. This is listing on China's STAR Market, not Hong Kong, not the US, which makes access extremely difficult for international investors. The risks are real. The US House Select Committee on the CCP has formally requested Unitree be blacklisted. Their robots appeared in PLA military exercises in 2024. Tariffs have already nearly tripled the US price of the G1. $TSLA Optimus is targeting sub $20,000 pricing with automotive scale manufacturing backed by $NVDA compute. If they execute, the price advantage shrinks fast. But this is still the only profitable pure play humanoid robotics company in the world growing at 335% a year, valued at a fraction of its loss making peers. Goldman projects the humanoid market at $38 billion by 2035. Morgan Stanley goes to $5 trillion by 2050. Unitree currently holds the largest market share of any humanoid manufacturer on the planet. Full breakdown coming soon. $TSLA $NVDA

KawzInvests

74,078 次观看 • 5 个月前

Every Wall Street giant that owns an AI data center is suddenly looking for a buyer. And NONE of them want to be the last one holding it. Three of them made their move in the last two weeks: Vantage Data Centers is exploring an exit. Its owners, Silver Lake and DigitalBridge, are weighing a listing at around $100 billion, or a sale, or a stake sale. It would be the largest data center IPO ever done. Three days earlier, CyrusOne started the same process. KKR and Global Infrastructure Partners met Goldman Sachs and Morgan Stanley, and the banks pitched for roles on a listing that could come as early as 2027. Last month, Switch hired Goldman and JPMorgan to take it public at close to $80 billion including debt, possibly by the fourth quarter. Three different companies moved inside the same 14 days, and the same handful of investment banks took every call. And these are the exact same firms that BOUGHT these companies off the public market four years ago. Between June 2021 and early 2022, private equity took the data center industry private. Blackstone bought QTS. KKR and Global Infrastructure Partners took CyrusOne private in a deal worth about $15 billion. DigitalBridge and IFM took Switch private for about $11 billion. Together those deals ran past $35 billion. By 2023 there were only two pure-play data center companies left on the public market. The logic at the time was that data centers burn cash for years before they pay, and public shareholders hate that. But private money was patient, and private money could wait. Four years later, the AI boom arrived and every one of those buildings became a gold mine. So follow this: Switch went private at about $11 billion in 2022. Its owners now want close to $80 billion for it. That is roughly 7x, in four years, on the same buildings. And DigitalBridge sits on both sides of this. It owns a piece of Vantage and it took Switch private. It is now looking for the door on BOTH. The question now is who is supposed to buy. There is no bigger private buyer left to sell to. These are already the largest infrastructure funds on Earth, and the price tags now run to $100 billion. The only pocket deep enough is the public market, which means anyone with a brokerage account or an index fund. The people who bought low from the public are now organizing to sell high back to the public. And they are doing it while telling everyone the buildout is just getting started. KKR raised a record $19.2 billion for its newest infrastructure fund this month, and in June launched a separate company with over $10 billion committed to finance more construction. So one hand raises fresh billions to build more data centers, and the other hand sells the finished ones to whoever will take them. None of this proves anyone thinks the boom is ending. Selling into strength is what these firms are paid to do, and every one of these deals is early stage and might never happen. But the timing tells you something: The most sophisticated infrastructure investors alive spent four years accumulating these assets in private, and all decided in the same two weeks that now is the moment to find someone else to own them. Four years ago these firms decided the public market was too impatient to own data centers. Now they want the public market to own them again, at 7x the price. Quite suspicious.
1:52

Every Wall Street giant that owns an AI data center is suddenly looking for a buyer. And NONE of them want to be the last one holding it. Three of them made their move in the last two weeks: Vantage Data Centers is exploring an exit. Its owners, Silver Lake and DigitalBridge, are weighing a listing at around $100 billion, or a sale, or a stake sale. It would be the largest data center IPO ever done. Three days earlier, CyrusOne started the same process. KKR and Global Infrastructure Partners met Goldman Sachs and Morgan Stanley, and the banks pitched for roles on a listing that could come as early as 2027. Last month, Switch hired Goldman and JPMorgan to take it public at close to $80 billion including debt, possibly by the fourth quarter. Three different companies moved inside the same 14 days, and the same handful of investment banks took every call. And these are the exact same firms that BOUGHT these companies off the public market four years ago. Between June 2021 and early 2022, private equity took the data center industry private. Blackstone bought QTS. KKR and Global Infrastructure Partners took CyrusOne private in a deal worth about $15 billion. DigitalBridge and IFM took Switch private for about $11 billion. Together those deals ran past $35 billion. By 2023 there were only two pure-play data center companies left on the public market. The logic at the time was that data centers burn cash for years before they pay, and public shareholders hate that. But private money was patient, and private money could wait. Four years later, the AI boom arrived and every one of those buildings became a gold mine. So follow this: Switch went private at about $11 billion in 2022. Its owners now want close to $80 billion for it. That is roughly 7x, in four years, on the same buildings. And DigitalBridge sits on both sides of this. It owns a piece of Vantage and it took Switch private. It is now looking for the door on BOTH. The question now is who is supposed to buy. There is no bigger private buyer left to sell to. These are already the largest infrastructure funds on Earth, and the price tags now run to $100 billion. The only pocket deep enough is the public market, which means anyone with a brokerage account or an index fund. The people who bought low from the public are now organizing to sell high back to the public. And they are doing it while telling everyone the buildout is just getting started. KKR raised a record $19.2 billion for its newest infrastructure fund this month, and in June launched a separate company with over $10 billion committed to finance more construction. So one hand raises fresh billions to build more data centers, and the other hand sells the finished ones to whoever will take them. None of this proves anyone thinks the boom is ending. Selling into strength is what these firms are paid to do, and every one of these deals is early stage and might never happen. But the timing tells you something: The most sophisticated infrastructure investors alive spent four years accumulating these assets in private, and all decided in the same two weeks that now is the moment to find someone else to own them. Four years ago these firms decided the public market was too impatient to own data centers. Now they want the public market to own them again, at 7x the price. Quite suspicious.

Ricardo

71,148 次观看 • 1 个月前

Matt Damon explains why a $25,000,000 movie must make $100,000,000 to break even after DVDs became obsolete “The economics of trying to make a kind of mid range drama now are just so terrible. If you have a $25 million movie, you have to put that much into print and advertising, so that’s $50 million. You split your money between the distributor and the exhibitor, so the people who own the movie theaters get half of the money. That means if your cost is now at $50 million, you need to make $100 million just to break even” “That’s why Behind the Candelabra, we did it at HBO because we couldn’t get funding from any studio. They said, ‘We love you guys. We believe in you. We think the movie is going to be great, but would I bet that this is going to make $100 million. Would I gamble $50 million on that assumption. It’s just too much’” “That’s because the DVD market is gone, and technology just erased the DVD. That was a huge revenue stream for all of us because you could afford for your movie to come out in a theater and not make everything back. It could make it partially back, and you’d know the DVD was coming to help fill the gap”
1:06

Matt Damon explains why a $25,000,000 movie must make $100,000,000 to break even after DVDs became obsolete “The economics of trying to make a kind of mid range drama now are just so terrible. If you have a $25 million movie, you have to put that much into print and advertising, so that’s $50 million. You split your money between the distributor and the exhibitor, so the people who own the movie theaters get half of the money. That means if your cost is now at $50 million, you need to make $100 million just to break even” “That’s why Behind the Candelabra, we did it at HBO because we couldn’t get funding from any studio. They said, ‘We love you guys. We believe in you. We think the movie is going to be great, but would I bet that this is going to make $100 million. Would I gamble $50 million on that assumption. It’s just too much’” “That’s because the DVD market is gone, and technology just erased the DVD. That was a huge revenue stream for all of us because you could afford for your movie to come out in a theater and not make everything back. It could make it partially back, and you’d know the DVD was coming to help fill the gap”

dank

80,767 次观看 • 1 个月前

Ladies and gentlemen, something happened in Tuapse, and we received a video message from the locals. This video is about humanity. This video is about you and me... Why are you, Ukrainians, so brainwashed, angry, and unsympathetic to the grief of the people of Tuapse... And aren’t you ashamed to be like that... But let's be serious. It's 2026, something happened, and instead of Z-festivals, dear Russians were recording videos saying they are just pawns, that they are not guilty, that they sympathize with us, and that they never wanted any of this. And you shouldn't hate them, they are... good. They just want peace, for unicorns to run across the rainbow, and instead of rockets, to launch confetti 🥰 And along with you, madam, there are another 1 million mercenaries fighting in Ukraine, 2.5 million regular scumbags, 2.5 million weaklings, more than 4 million scumbags working in defense industries, 3 million scumbags in the government apparatus. And tens of millions more scumbags to some degree supporting the Kremlin regime. And each. Each of these people deserves liquidation. Because you are a plague, a contagion that must be burned out. Probably if a cancer cell could speak during therapy, it would say something like this: I just exist. I don’t want to kill you. I’m very sorry that you suffer. I wouldn’t want you to die. Can you negotiate with a cancer cell? Can you expect a cancer cell to pity you? Can you expect cancer to retreat on its own? ... No. The so-called Russian Federation is cancer. The enraged, feral Russian society is cancer. Cancer must be burned out, cut out. Every cell must be destroyed. Because cancer will never retreat on its own. If you disagree with the war, if you don’t support your government — leave the country. Until then, each of you is an implicit accomplice and a terrorist. And there’s no whining. Sorry for the long read, it can’t be shorter here.
0:11

Ladies and gentlemen, something happened in Tuapse, and we received a video message from the locals. This video is about humanity. This video is about you and me... Why are you, Ukrainians, so brainwashed, angry, and unsympathetic to the grief of the people of Tuapse... And aren’t you ashamed to be like that... But let's be serious. It's 2026, something happened, and instead of Z-festivals, dear Russians were recording videos saying they are just pawns, that they are not guilty, that they sympathize with us, and that they never wanted any of this. And you shouldn't hate them, they are... good. They just want peace, for unicorns to run across the rainbow, and instead of rockets, to launch confetti 🥰 And along with you, madam, there are another 1 million mercenaries fighting in Ukraine, 2.5 million regular scumbags, 2.5 million weaklings, more than 4 million scumbags working in defense industries, 3 million scumbags in the government apparatus. And tens of millions more scumbags to some degree supporting the Kremlin regime. And each. Each of these people deserves liquidation. Because you are a plague, a contagion that must be burned out. Probably if a cancer cell could speak during therapy, it would say something like this: I just exist. I don’t want to kill you. I’m very sorry that you suffer. I wouldn’t want you to die. Can you negotiate with a cancer cell? Can you expect a cancer cell to pity you? Can you expect cancer to retreat on its own? ... No. The so-called Russian Federation is cancer. The enraged, feral Russian society is cancer. Cancer must be burned out, cut out. Every cell must be destroyed. Because cancer will never retreat on its own. If you disagree with the war, if you don’t support your government — leave the country. Until then, each of you is an implicit accomplice and a terrorist. And there’s no whining. Sorry for the long read, it can’t be shorter here.

Exilenova+

31,086 次观看 • 4 个月前

The US is about to charge $30 million per tanker to cross the Strait of Hormuz. Trump just declared the US the "Guardian of the Hormuz Strait" and said it will take a 20% cut on all cargo passing through. Here is what that actually means. A fully loaded supertanker carries about 2 million barrels of oil. At $75 a barrel, that cargo is worth roughly $150 million. A 20% fee on that is $30 million. Per ship. Per crossing. Now compare that to what Iran was charging. Iran's toll has been running at $1.5 million to $2 million per vessel. On a $150 million cargo, that is about 1.3%. Trump called that toll unacceptable. His replacement is roughly 15 times more expensive. The scale of this is what nobody is talking about. Before the war, 20.3 million barrels of oil crossed Hormuz every single day. At $75 oil, that is $1.52 billion of crude moving through the strait daily. A 20% cut on that comes to roughly $304 million a day. That is about $111 billion a year. For comparison, Iran's entire toll system was projected to earn $1 billion to $2 billion a year at best. The US plan would collect more than 50 times that. There is no precedent for this anywhere in global trade. The Suez Canal charges roughly $300,000 to $700,000 per vessel. The Panama Canal is similar. Both are man-made canals that countries built and maintain. Hormuz is a natural waterway. Under international law, ships have a right of transit passage through it. That is the exact legal argument the US used against Iran's toll. And the cost does not land on the US. It lands on Saudi Arabia, the UAE, Qatar, Kuwait, and Iraq, who ship the oil. And on China, India, Japan, and South Korea, who buy it. A $30 million fee per tanker works out to $15 per barrel. That gets passed straight into the price of crude. Oil is already up over 4% today. The strait that was supposed to reopen and lower prices is now being turned into the most expensive stretch of water on earth.
0:28

The US is about to charge $30 million per tanker to cross the Strait of Hormuz. Trump just declared the US the "Guardian of the Hormuz Strait" and said it will take a 20% cut on all cargo passing through. Here is what that actually means. A fully loaded supertanker carries about 2 million barrels of oil. At $75 a barrel, that cargo is worth roughly $150 million. A 20% fee on that is $30 million. Per ship. Per crossing. Now compare that to what Iran was charging. Iran's toll has been running at $1.5 million to $2 million per vessel. On a $150 million cargo, that is about 1.3%. Trump called that toll unacceptable. His replacement is roughly 15 times more expensive. The scale of this is what nobody is talking about. Before the war, 20.3 million barrels of oil crossed Hormuz every single day. At $75 oil, that is $1.52 billion of crude moving through the strait daily. A 20% cut on that comes to roughly $304 million a day. That is about $111 billion a year. For comparison, Iran's entire toll system was projected to earn $1 billion to $2 billion a year at best. The US plan would collect more than 50 times that. There is no precedent for this anywhere in global trade. The Suez Canal charges roughly $300,000 to $700,000 per vessel. The Panama Canal is similar. Both are man-made canals that countries built and maintain. Hormuz is a natural waterway. Under international law, ships have a right of transit passage through it. That is the exact legal argument the US used against Iran's toll. And the cost does not land on the US. It lands on Saudi Arabia, the UAE, Qatar, Kuwait, and Iraq, who ship the oil. And on China, India, Japan, and South Korea, who buy it. A $30 million fee per tanker works out to $15 per barrel. That gets passed straight into the price of crude. Oil is already up over 4% today. The strait that was supposed to reopen and lower prices is now being turned into the most expensive stretch of water on earth.

The Macro Paper

63,688 次观看 • 2 个月前

Mark my words, Nebius will be the first Trillion dollar Neo-cloud company and here is why (Save this). Roman Chernin, CEO of Nebius just said on 20VC that Nebius raised prices and demand didn't move. When a company can raise prices and still have more demand than supply, that's the opportunity. Chernin also explained why he is deliberately not charging the maximum. As AI shifts from training, a one time cost to inference, which is the ongoing cost of serving every user and every query, compute pricing becomes the cost structure of the entire AI economy. If Nebius prices customers out, those customers cannot grow, and Nebius cannot grow with them. That is the compounding flywheel built directly into the revenue model. The numbers are already confirming it. Q1 2026 revenue came in at $399 million, up 684% year over year. The AI cloud segment grew 840% and represented 98% of total revenue. Adjusted EBITDA flipped positive to $129.5 million. And Nebius signed a long-term agreement with Meta worth up to $27 billion over five years, a hyperscaler outsourcing its own AI compute stack to a neocloud, which tells you that even companies with $50 billion capex budgets cannot build fast enough. Goldman Sachs says the consensus is underestimating 2027 hyperscaler capex by $500 billion. Every dollar hyperscalers cannot provision themselves flows to neoclouds like Nebius. As that gap widens, Nebius captures the overflow with 3 gigawatts of contracted power already secured and a CEO who just told you raising prices did not dent demand. Our subscribers are already up massively on Nebius and come join Milk Road Pro for our full breakdown, how to size Nebius against the broader neocloud opportunity, and our full AI thesis. Link below!
0:47

Mark my words, Nebius will be the first Trillion dollar Neo-cloud company and here is why (Save this). Roman Chernin, CEO of Nebius just said on 20VC that Nebius raised prices and demand didn't move. When a company can raise prices and still have more demand than supply, that's the opportunity. Chernin also explained why he is deliberately not charging the maximum. As AI shifts from training, a one time cost to inference, which is the ongoing cost of serving every user and every query, compute pricing becomes the cost structure of the entire AI economy. If Nebius prices customers out, those customers cannot grow, and Nebius cannot grow with them. That is the compounding flywheel built directly into the revenue model. The numbers are already confirming it. Q1 2026 revenue came in at $399 million, up 684% year over year. The AI cloud segment grew 840% and represented 98% of total revenue. Adjusted EBITDA flipped positive to $129.5 million. And Nebius signed a long-term agreement with Meta worth up to $27 billion over five years, a hyperscaler outsourcing its own AI compute stack to a neocloud, which tells you that even companies with $50 billion capex budgets cannot build fast enough. Goldman Sachs says the consensus is underestimating 2027 hyperscaler capex by $500 billion. Every dollar hyperscalers cannot provision themselves flows to neoclouds like Nebius. As that gap widens, Nebius captures the overflow with 3 gigawatts of contracted power already secured and a CEO who just told you raising prices did not dent demand. Our subscribers are already up massively on Nebius and come join Milk Road Pro for our full breakdown, how to size Nebius against the broader neocloud opportunity, and our full AI thesis. Link below!

Milk Road AI

15,677 次观看 • 3 个月前

David Friedberg: Starlink alone could be a trillion-dollar company "Starlink alone could be generating on the order of $40 billion of revenue top line, with a huge amount of that flowing to free cash. That could be $30 billion in free cash flow within the year." "That alone provides the cash flow to fund much of what Elon is doing. And if you just put a 30x multiple on that, which I think you can, because these subscription businesses have very high renewal rates and very low customer acquisition costs (CAC)." "The Starlink business alone could be a trillion-dollar market cap within 18 months. That alone funds all the rest of this." "You can see it in AT&T, Verizon, HughesNet, and Viasat... That market got decimated by Starlink." "If he launches the handset thing, that subscriber growth is going to go. Right now, he's adding 2 million subscribers on the consumer side per quarter. You could see that going to 4 to 5 million a quarter. You could actually see an acceleration. There are 400 million mobile subs just in the U.S." "And the rest of it is like, is Elon going to do well with the excess capital that's spinning out of Starlink? Well, I don't know who else I would give it to, considering what he's doing with Starship, AI compute, and the Terafab." Elon Musk david friedberg The All-In Podcast
3:19

David Friedberg: Starlink alone could be a trillion-dollar company "Starlink alone could be generating on the order of $40 billion of revenue top line, with a huge amount of that flowing to free cash. That could be $30 billion in free cash flow within the year." "That alone provides the cash flow to fund much of what Elon is doing. And if you just put a 30x multiple on that, which I think you can, because these subscription businesses have very high renewal rates and very low customer acquisition costs (CAC)." "The Starlink business alone could be a trillion-dollar market cap within 18 months. That alone funds all the rest of this." "You can see it in AT&T, Verizon, HughesNet, and Viasat... That market got decimated by Starlink." "If he launches the handset thing, that subscriber growth is going to go. Right now, he's adding 2 million subscribers on the consumer side per quarter. You could see that going to 4 to 5 million a quarter. You could actually see an acceleration. There are 400 million mobile subs just in the U.S." "And the rest of it is like, is Elon going to do well with the excess capital that's spinning out of Starlink? Well, I don't know who else I would give it to, considering what he's doing with Starship, AI compute, and the Terafab." Elon Musk david friedberg The All-In Podcast

KanekoaTheGreat

473,312 次观看 • 1 个月前