Podcast Notes: Samsung Earns 650 Million Daily Surpassing NVIDIA, Yet Stock Price Plunges 10%, This Is "Selling News" Rather Than Storage Doomsday
TechFlow Selected TechFlow Selected
Podcast Notes: Samsung Earns 650 Million Daily Surpassing NVIDIA, Yet Stock Price Plunges 10%, This Is "Selling News" Rather Than Storage Doomsday
For Every $100 of Memory Sold, SK Hynix Takes $72 in Profit—Are the Kings of Storage Being Misread?
Navigating Web3 tides with focused insightsOrganized & Compiled: TechFlow
Guest: EJ, Co-host of Limitless Podcast
Host: Josh, Limitless Podcast
Podcast Source: Limitless Podcast (formerly Bankless Channel)
Original Title: The AI Trade Everyone's Getting Wrong
Air Date: July 9, 2026
Key Takeaways
Samsung's latest quarterly profit was $58.5 billion, exceeding NVIDIA's $53 billion for the same period, with over 94% of that profit coming from one division: AI memory. Only three companies worldwide can produce HBM (High Bandwidth Memory), two in South Korea (Samsung and SK Hynix) and one in the US (Micron). Every AI inference requires rereading the entire model weights; memory demand is 10 to 20 times that of consumer products, and it doubles with every generation of models. Meanwhile, 1 GB of HBM consumes capacity equivalent to 4 GB of standard DRAM, directly squeezing out memory supply for phones and computers, resulting in price increases across all Apple products.
The contradiction lies in this: memory companies are hitting record profits and prices are still rising, yet stock prices have collectively fallen into a bear market. Samsung fell 9% on earnings day, SK Hynix fell 15%. The two hosts believe this is a combination of a "sell the news" event and cyclical panic, with no change in fundamentals. SK Hynix listed on NASDAQ in the form of ADRs, raising approximately $30 billion, with 4x oversubscription, and Leopold Aschenbrenner participated as a seed investor.
Highlights Summary
Memory Monopoly Profit Margins Crush Everything
"Samsung's gross margin is 52%, SK Hynix is 72%. For every $100 of memory sold, $72 goes directly onto the balance sheet. For comparison, Apple's hardware gross margin is about 30%."
"Samsung memory division employees received year-end bonuses worth 6 times their annual salary. South Korea's luxury goods market sales tripled over the past 4 months."
AI is a Memory Black Hole
"AI is basically a memory black hole, more exaggerated than any consumer or enterprise technology wave we've seen."
"Every time you submit a prompt, whether it's ChatGPT or Claude, it has to reread the entire model weights. Every single time."
"Each new model's demand for memory is 10 to 20 times that of the previous generation. With 15 trillion parameter, 20 trillion parameter models, this demand will only go up."
The Price Surge Isn't Over Yet
"Q1 rose 90%, Q2 rose another 50 to 60%, Q3 will rise another 20%. Samsung earned more money in one year than in the past 40 years combined."
Profits Hit Records, Stock Prices Fall into Bear Market
"Samsung beat expectations on earnings day, single-quarter profit exceeded NVIDIA. Then the stock price fell 9%. SK Hynix fell 15%."
"Meta hinted at curbing AI capital expenditure. Honestly, they haven't made anything useful anyway."
"If you bought a month ago, it looks pretty stupid now. But 6 to 24 months later, these companies are still very strong businesses."
Main Text
Samsung: The Misunderstood King of AI Profits
Josh: Samsung has just become the most profitable company globally, but almost no one knows what they actually make money from. I used to think of Samsung as smartphones and computers, but in reality, 96% of their profit comes from one division, which is memory. There are only three major suppliers in the memory market, two of which happen to be in South Korea. The two most profitable companies in the world are in the same country, what exactly is going on? EJ, how did Samsung do it?
EJ: When I was young, I only knew Samsung sold phones. Over the past year and a half, I kept hearing this name tied to AI, and I was also curious what this company was actually doing. Let's look at some numbers first. Samsung Q2 quarterly profit was $58.5 billion, analyst expectations were $55 billion, beating by $3 billion. More importantly, they crushed NVIDIA, NVIDIA only had $53 billion for the same period. One year ago, Samsung Q2 2025 profit was only $3.4 billion. From $3.4 billion to $58.5 billion in one year, what happened in between?
The answer is HBM, High Bandwidth Memory. Samsung is currently the second largest HBM supplier globally. Why is HBM so valuable? Because every GPU sold by NVIDIA, every TPU manufactured by Google, any AI chip requires a large amount of memory. Note the word "large," the amount of memory needed for consumer devices (laptops, computers) is predictable, but AI memory demand is exponential.
Samsung earns $650 million per day, $27 million per hour, $7,500 per second. This is more than the most valuable company globally earns. But Samsung is not the most valuable company globally, which is very interesting.
The Big Three Memory Players and HBM Monopoly
Josh: Computers need two things: a workbench (memory) and a filing cabinet (storage). There are only three companies globally that can make these things.
EJ: Correct. Three types of memory, each with its own use. DRAM is the DDR5 in your computer, the workbench, temporary RAM, cleared when powered off. NAND is your SSD, the flash memory in iPhones, the filing cabinet, slower but cheaper. Then there is HBM, High Bandwidth Memory, this is everything in the AI era.
HBM manufacturing is extremely complex. DRAM chips are stacked 12 to 16 layers high like skyscrapers, using extremely precise processes. SK Hynix occupies 60% of the entire HBM market. Any AI chip produced, most likely uses their memory. The monopoly pattern of the three companies has formed, and profit margins have expanded accordingly.
Josh: Here is a very intuitive number. 1 GB of HBM consumes wafer capacity equivalent to 4 GB of standard DRAM. That means, for every wafer shifted to AI memory, 4 times the phone and computer memory capacity disappears. Apple raised prices across the board for the first time, MacBook Air from $1100 to $1300, MacBook Pro from $1700 to $2000, Mac Studio from $4000 to $5300. Because people who buy Mac Studio want to run local inference, memory demand is huge, and there just isn't enough memory.
AI is a Memory Black Hole
EJ: If I had to explain in one sentence why everyone is so bullish on these companies: AI is basically a memory black hole, more exaggerated than any technology wave we've seen.
You write a prompt and submit it, whether it's ChatGPT or Claude, it has to reread the entire model weights every time. What are model weights? They are the parameters those companies spent billions of dollars training. Every inference rereads them once. Moreover, each new model's demand for memory is 10 to 20 times that of the previous generation. 15 trillion parameter, 20 trillion parameter models, demand will only continue to push upwards.
But this isn't the biggest part. Your chatbot remembers what you said last time, remembers your information across conversations, this temporary storage goes through the NAND flash memory part, demand is equally huge. So all three types of memory demand are simultaneously exploded.
Some say this is a bubble, it will burst. Historically this makes sense, the memory industry has always had cycles. Three years ago SK Hynix was almost acquired by Micron, that was the absolute bottom of the cycle. They invested in HBM, didn't know if they could sell it, Micron almost bought them. In the end they didn't sell, doubled down on HBM, and now have become the company with the highest market cap in South Korea.
Price Surge: From 90% to 20% Growth Steps
EJ: Let's look at prices. SK Hynix and Samsung pricing over the past 6 months: Q1 rose 90%. The memory in your usual phone and computer, suddenly prices doubled, just because one component rose nearly 100%. Q2 rose another 50 to 60%. Samsung Q3 will rise another 20%.
Samsung earned more money in one year than in the past 40 years combined, 19 times that of the same period last year.
In terms of gross margin, grocery stores earn $3 for every $100 of goods sold, car companies earn $7, Apple hardware earns $30. Samsung is 52, SK Hynix is 72. For every $100 of memory sold, $72 goes directly into the account. Samsung memory employees received year-end bonuses worth 6 times their annual salary. South Korea's luxury goods market tripled over the past 4 months. Someone in Taiwan borrowed a $60,000 high-interest loan from a bank to buy TSMC. Some crazy things are happening in the Asian AI market.
Josh: This increase is too fierce. Consumers are already bearing the cost. 32GB memory sticks are 2 to 3 times more expensive than last year. Building a computer, one-third of the cost is memory. These prices have already transmitted to the real consumer market. The question is: How long can this last? Can they keep rising?
EJ: This depends on one variable: Will the number of people using AI continue to increase? This is the only proxy indicator to measure whether memory demand will continue to grow. If you think everyone will run multiple AI agents in the future, using AI for work and life, then memory demand is exponential.
What about the supply side? New wafer fabs won't come online until 2030. These fabs are designed extremely precisely, impossible to flood supply in the short term. Demand is 3 to 5 times faster than supply, supply constrained for the next few years. On the China side, CXMT is making similar DRAM and HBM, but their capacity is all eaten up by Chinese local AI labs. Apple wanted to find alternative suppliers from China, no stock either.
There is another argument: What if new model architectures appear that don't need so much memory? I think the logic is exactly the opposite. Memory becomes cheaper, more scenarios can run, more AI agents are deployed, overall economic output is higher. Demand for memory is actually greater.
Contradiction: Profits Hit Records, Stock Prices Fall into Bear Market
Josh: We have always been bullish on memory, Micron has risen 150% since recommended at the end of last year. But recently all memory stocks have fallen more than 20% from highs. Technical bear market. Samsung beat expectations on earnings day, single-quarter profit exceeded NVIDIA, then fell 9%. SK Hynix fell 15%. Profits are hitting records, prices are still rising, but the market is saying "slow down".
What might have scared the market is Meta hinting at curbing AI capital expenditure. But honestly they haven't made anything useful anyway. This is the current strange node: companies say "we are good, profits hit records, demand is strong", the market says "wait, risen too much, there are unknown risks". What do you think?
EJ: My judgment is simple: this is sell the news. The quarterly earnings season just ended, global funds have heavy positions in these stocks, want to wait for a good price to re-enter. If you think I'm being stubborn, understandable. But if you are a long-term AI investor, memory is a rigid demand, and only these three companies can produce it, this pattern won't change in the short term.
What people worry about is the cycle. 2017 to 2018 the last memory super cycle, Micron's PE reached 4 to 5 times, then fell 60%, although profits were still rising. History is indeed repeating, but the difference this time is: Last time was phone-driven, you could predict the demand ceiling. This time is AI-driven, the demand ceiling is unseen.
There is also an interesting historical detail. At the bottom of the last cycle, when memory prices were cheapest, who was desperately suppressing prices to hoard memory? Apple. Apple had pricing power over Samsung and SK Hynix at that time, forcing them to supply DRAM and HBM at the lowest price. Now the tables have turned, Samsung and SK Hynix are doing what Apple did back then. In an efficient market, this is normal.
SK Hynix Lands on NASDAQ
Josh: Next there is a heavyweight test. SK Hynix is a South Korean company, but listed on NASDAQ in the form of ADRs on July 10, raising approximately $30 billion. This is a critical moment for the memory industry, how the US market prices SK Hynix determines the next phase of this memory trade. EJ, will you participate?
EJ: Short answer: Yes. I am bullish on memory. I hold Micron, also hold DRAM ETF (a basket of memory companies). If you are a US citizen, buying South Korean stocks directly is inconvenient, but can allocate through these basket products. I waited a long time for SK Hynix to list in the US, will buy.
This IPO is currently rumored to be 4x oversubscribed. Institutions, pension funds, retail investors are all grabbing. Being able to get this scale indicates institutions have done deep research on the memory track, planning to hold long-term.
Josh: Guess who $2 to $3 billion of this $30 billion comes from? Leopold Aschenbrenner. He is back again. Last time we did an episode analyzing his holdings, some questioned his logic on shorting NVIDIA. As a result, NVIDIA has fallen 20% from when we recorded that episode to now. Now he participates in SK Hynix's IPO as a seed investor again. You can question his judgment, but he indeed hasn't missed.
Long-term Perspective: Super Cycle or Bubble?
Josh: Let's align on conclusions. Long term, memory demand has no ceiling, only these three companies can make it, new entrants cannot form meaningful competition in the short term. Short term, people are afraid, because of historical cycle shadows, risen too much too fast. This makes complete sense. What you bought a month ago, looks pretty stupid now. But 6 to 24 months later, the fundamentals of these companies won't change.
EJ: Correct. There will eventually be a cycle top, that must be when wafer capacity starts to be in oversupply. But supply won't come up until 2030, demand is 3 to 5 times supply. Before that, margin and profit rates will continue to expand.
I am more curious what listeners think. AI memory is not the sexiest topic, but it is becoming the most important component in the AI infrastructure capital expenditure race. It is the source pushing up all consumer goods prices. Do you think we are too optimistic? Can profit margins continue to rise? Or is this a bubble that will eventually burst, and are Josh and I marking the top?
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
深潮TechFlow
