The numbers landed like a wrench in the machine. On July 7, 2024, Samsung Electronics announced its preliminary earnings for the second quarter: operating profit surged to approximately 10.9 trillion Korean won, an eighteen-fold increase from the 0.67 trillion recorded in the same period last year. Revenue hit 75 trillion won, a 20% year-over-year gain. The market blinked. Analysts scrambled to revise models. But I wasn't surprised; I saw the glass cracking weeks earlier.
For months, I had been watching the silent signals—the microscopic shifts in global liquidity that precede the noise. Samsung, the world's largest memory chip maker, was not just reporting a quarterly earnings beat. It was issuing a macro confession: we are in a new cycle, and it is not about smartphones.
My eye is on the horizon, not the hourly candle.
A Necessary Pruning
The bust was not an end, but a necessary pruning. For nearly two years, the semiconductor industry endured a brutal downcycle. Prices of DRAM and NAND flash memory collapsed. Inventories piled up like digital debris. Samsung, for all its IDM might, saw its operating profit spiral from the 2021 peak of 15.6 trillion won to a terrifying 4.7 trillion in Q1 2023, then to a mere 0.6 trillion in Q2 2023. It was the kind of ice-cold contraction that kills companies with weaker balance sheets.
Yet the survivors emerge leaner, stronger, and smarter. Samsung’s Q2 2024 numbers—operating profit of 10.9 trillion won, revenue of 75 trillion won—are not a cyclical bounce; they are a structural metamorphosis. The company’s semiconductor division alone likely generated over 8 trillion won in profit, driven by the explosive demand for High Bandwidth Memory (HBM) and DDR5 from AI data centers. The consumer electronics and mobile businesses, once the profit engines, now play a supporting role. This is a tectonic shift.
To understand this, we must map the global liquidity flows. The Q2 earnings reflect three macro forces converging: the AI hardware super-cycle, the forced rebuilding of supply chains due to US-China decoupling, and the end of the inventory correction. The crypto world would be wise to watch.
The Liquidity Lens
Crypto is a macro asset, not a rebel. When the Fed tightens, risk assets fall. When liquidity flows into AI chips, capital escapes from meme coins. The story of Samsung’s profit surge is the story of where global capital is actually going: into physical, high-compute infrastructure. The servers that train GPT-5 and power Bitcoin mining use the same high-bandwidth memory chips. The demand for HBM is another proxy for the demand for computational sovereignty. As I wrote in a recent internal memo for our fund: “The new class of liquidity flows is into compute cycles, not token cycles.”
But there is a profound nuance here. The crypto market’s current sideways consolidation is often interpreted as weakness. I see it as a foundational phase. The chop is for positioning. While retail attention flits between AI tokens and Layer2 narratives, the structural money is flowing to the proven infrastructure: the chips that enable block production, verification, and storage.
Samsung’s three hundred and sixty degree return to profitability is not just a story of smartphones; it is about AI data center spending. Hyperscale cloud providers—who are also the custodians of many blockchain nodes—are ordering HBM3E chips in bulk. This is a direct injection of real compute into the supply chain of the digital future. The cycle is not about sentiment; it is about capital formation.
Based on my experience auditing liquidity cycles during the 2019-2020 bear market, I identified three technical signals in Samsung's Q2 data that matter for crypto investors.
Signal One: The Memory Price Inflection
The most direct trigger is the price of memory chips. After a 50%+ decline from peak to trough, DRAM and NAND prices began to stabilize in early 2024 and then surged in Q2. Samsung’s data shows that DRAM bit shipments grew by a mid-single-digit percentage sequentially, but pricing improved by over 15%. This is a textbook signal that the inventory cycle has flipped from destocking to restocking. For crypto, this means the cost of hardware for mining and decentralized storage will rise. If you are farming Filecoin or building a GPU-based DePIN (Decentralized Physical Infrastructure Network), your cost basis just increased. This could compress margins in those sectors, leading to a rationalization of marginal participants. Winter clears the weak hands.
Signal Two: HBM as a Liquidity Proxy
HBM3E is not just a memory chip; it is a measure of AI training intensity. Samsung’s HBM shipments in Q2 were likely up over 50% quarter-over-quarter, driven by NVIDIA’s H200 and B100 chips. This is a direct correlation to the scale of AI workload growth. For the crypto ecosystem, which is increasingly integrating AI (e.g., decentralized compute networks like Render or Akash), rising HBM production indicates both opportunity and threat. Opportunity: the hardware backbone is being built. Threat: centralized AI infrastructure is hardening its lead over decentralized alternatives.
In my 2024 quantitative model for the fund, we included a “HBM Supply Index” as a leading indicator for the viability of decentralized AI projects. The recent surge in Samsung’s output suggests that the supply of high-performance memory is accelerating, which lowers the barrier for new entrants but also shifts the bottleneck from hardware to software. The next phase of the AI-blockchain integration will be about algorithms, not chips.
Signal Three: The CAPEX Rollout
Samsung’s semiconductor capital expenditure in Q2 was approximately 13 trillion won, an increase of 15% from the previous quarter. This is not reckless spending; it is strategic positioning. The company is expanding its HBM capacity at the Pyeongtaek campus and building a new advanced packaging plant in Taylor, Texas, to get closer to the US AI market. This is a direct response to geopolitical risk.
For the crypto world, this trend of “fawnshoring”—moving critical supply chains to friendly jurisdictions—mirrors the migration of crypto miners and node operators out of China and into North America after the 2021 ban. Samsung’s CAPEX is a macro map of where physical infrastructure will be located. If your DePIN project depends on low-cost electricity in China or hardware supply from Southeast Asia, you are exposed to the same geopolitical risks that Samsung is now hedging against.
The Decoupling Thesis
Now, the contrarian angle. Most analysts celebrate Samsung’s profit surge as a validation of the AI trade, but I see a deeper, more uncomfortable truth: this performance is a temporary consequence of geopolitical friction, not an expression of pure market demand.
Samsung’s Q2 victory lap is, in part, a reward for the US-China chip war. American sanctions on Chinese memory players like YMTC and CXMT have removed them as competitors in the advanced memory market. Samsung now has less competition and more pricing power. This is a “geopolitical dividend,” but it is not sustainable. A de-escalation in trade tensions or a reversal of export controls could suddenly flood the market with cheaper Chinese memory, compressing Samsung’s margins.
Furthermore, Samsung’s foundry business is still bleeding market share to TSMC. The 3nm GAA (Gate-All-Around) process, which Samsung once bet would leapfrog TSMC, has struggled with yield. This means that while Samsung is winning in memory, it is losing the battle for logic chips. In crypto, we understand that fragmentation is not scaling; it is slicing. The same applies to Samsung’s foundry division: it has too many products chasing too few high-volume clients.
This leads to my core decoupling thesis: the earnings of the memory business and the foundry business are decoupling from each other. In previous cycles, both moved together. Now, memory is booming while foundry is struggling. This implies that the company’s overall valuation cannot be justified by memory alone. For the crypto market, this is a cautionary tale. When a monolithic entity’s arms operate on different risk curves, the whole can be mispriced.
Finally, we must consider the regulatory pathway. The European Union’s MiCA regulation is creating a clearer framework for digital assets, while the US remains in a legal gray zone. Samsung’s physical expansion into the US is a hedge against future regulatory shifts. If the US enacts a comprehensive crypto regulatory framework that treats hardware as a security, projects reliant on high-compute chips could face new compliance costs. Samsung’s Texas plant is not just about making chips; it is about political insurance.
To quote one of my weekly briefs to institutional clients: “The regulatory clarity that Bitcoin ETF approval brought is the same clarity that is drawing Samsung’s CAPEX to American soil. Both are seeking the same anchor.”
Where We Stand
So, where does this leave the crypto investor in this sideways market? The chop is for positioning. The liquidity cycle is turning. Samsung’s earnings are a macro signal that real capital, not speculative capital, is entering the infrastructure layer. The next wave of crypto adoption will be powered by the same chips that are boosting Samsung’s profits. But the party will not be uniform.
DeFi protocols that rely on on-chain liquidity will need to understand that global liquidity is being diverted to hardware. The narrative of “Liquidity Fragmentation” is a convenient story VCs use to push new products, but the real fragmentation is between digital capital and physical capital. Samsung’s profit surge is proof that the latter is winning.
My own framework, shaped by the silence of the 2019 bust, tells me to watch the hardware cycles more closely than the Twitter sentiment. The bust was not an end, but a necessary pruning.
The question I leave you with is not “Should I buy Samsung stock?” but “Is your crypto portfolio positioned for a world where the most valuable assets are not tokens, but the compute cycles they depend on?