The SEC filing landed at 3:47 PM EST. Third Point LLC, Dan Loeb’s $20B hedge fund, had quietly shed its entire stake in Lam Research. The move was surgical—no public statement, no fanfare. Just a line item in a 13F that signals something far larger than a single portfolio adjustment. For those of us who track the invisible grid where value leaks out, this isn’t a semiconductor story. It’s a crypto mining hardware supply chain story. And it’s breaking now.
Lam Research is not a name you hear in crypto Twitter. It doesn’t mint tokens, run nodes, or validate blocks. But its plasma etch and atomic layer deposition machines are the silent architects of the silicon that powers Bitcoin ASICs, Ethereum validator GPUs, and the high-bandwidth memory (HBM) that feeds AI-driven trading algorithms. When a fund like Third Point—known for its event-driven, macro-aware plays—exits a position in Lam, it’s not a bet against etching technology. It’s a bet against the capital expenditure cycle that determines whether next-gen mining rigs get built, whether HBM3E production ramps, and whether the cost of compute for proof-of-work declines or spikes.
Context: Why Now, Why Lam? Lam Research is the third-largest wafer fab equipment (WFE) vendor globally, commanding ~20% of the $100B+ WFE market. Its core strength lies in high-aspect-ratio etching (critical for 3D NAND stacking beyond 200 layers) and deposition tools used in advanced packaging like TSV (through-silicon via) for HBM. Bitcoin ASIC manufacturers—Bitmain, MicroBT, Canaan—rely on foundries like TSMC and Samsung to produce chips at 5nm and 3nm nodes. Those foundries are Lam’s customers. Every new mining rig generation requires a new set of mask layers, etch steps, and deposition cycles. Lam’s equipment is the bottleneck.
But the relationship is indirect. Lam’s revenue is driven by foundry and memory maker capital expenditure, not by Bitcoin price directly. However, the correlation is tightening. In 2023-2024, the AI boom supercharged foundry spending on HBM and advanced logic, pushing Lam’s valuation to 30-35x trailing earnings—a historic premium. Third Point entered the position during the 2022 bear market when Lam traded at 15x. Now, with the stock up 150%+ from those lows, they are booking profits. The question is: what does this exit reveal about the next 12-18 months for crypto mining hardware?
Core: The Forensic Deconstruction of the Exit Let’s map the invisible grid. The core of Third Point’s thesis is likely a bet on the peaking of the global WFE cycle. Based on my forensic analysis of Lam’s order book and comparable filings, I see three structural signals embedded in this move.
First, China exposure is becoming a structural headwind, not a cyclical one. Lam’s China revenue dropped from ~29% of total in FY2021 to ~20-25% in FY2023 due to US export controls on advanced equipment. The CHIPS Act and the 2022/2023 BIS rules explicitly restrict the sale of equipment for 16nm/14nm and below logic, 18nm half-pitch DRAM, and 128+ layer 3D NAND. China’s domestic foundries—SMIC, Hua Hong—are pivoting to domestic tools from AMEC and Naura. Lam’s service revenue (high-margin, ~60%+ gross margin) remains, but new equipment sales to China are structurally capped. For crypto mining, this matters because China still hosts a significant portion of Bitcoin hashrate, and any new mining farm buildout in China will increasingly use domestic equipment for mature nodes. But the real impact is on the global supply chain: if China can’t upgrade its foundry capacity for advanced ASIC nodes, the next-gen mining rigs will be built exclusively in Taiwan, Korea, or the US, raising costs and lead times.
Second, HBM equipment demand is peaking faster than expected. Lam generates significant revenue from TSV etch and deposition tools used in HBM production. With SK Hynix, Samsung, and Micron racing to supply HBM3E for NVIDIA’s Blackwell and AMD’s MI300, Lam’s HBM-related revenue grew ~50% YoY in 2024. But the market is missing a key nuance: HBM technology is maturing. The transition from HBM3 to HBM4 will require hybrid bonding, which reduces the need for TSV etch and deposition per die. Lam’s tool intensity per wafer will decline. Third Point’s exit may reflect a front-running of this technology shift. For crypto miners, this is a double-edged sword: less HBM demand means more DRAM capacity available for general-purpose use, potentially lowering memory costs for mining rigs that use DRAM (like some FPGA-based setups). But the broader signal is that AI capital expenditure growth—the engine that lifted Lam—is decelerating.
Third, the operating leverage story is breaking. Lam’s gross margin has been steady at ~44-46%, but net margin is under pressure from export compliance costs and customer concentration risk. The US government’s “presumption of denial” policy on China licenses forces Lam to maintain a costly compliance infrastructure. Meanwhile, the top five foundries (TSMC, Samsung, Intel, SK Hynix, Micron) account for >70% of Lam’s revenue. Any single customer cutting capex (e.g., Intel pushing back its 18A ramp) would hit Lam disproportionately. Third Point, being an event-driven fund, likely modeled a scenario where global foundry capex peaks in 2025 and then declines 10-15% by 2027. That would compress Lam’s PE from 35x to 25x—a 30% downside even if earnings hold.
Contrarian Angle: The Unreported Blind Spot The conventional narrative is that Third Point is “selling the top” on AI hype. But the contrarian angle is this: the exit is actually a bet on the commoditization of crypto mining hardware. Let me explain.
Lam’s equipment is used to make the most advanced chips. But the Bitcoin mining industry is increasingly moving toward efficiency gains through scale, not node shrinks. The latest generation of ASICs (Antminer S21, Whatsminer M60) are already at 5nm/3nm. The next node (2nm) will require EUV double-patterning and new etch chemistries—Lam has tools for that, but the incremental hash rate gain per dollar spent is diminishing. If the network hash rate growth slows (as it has post-halving), the demand for new top-tier mining rigs may plateau. Third Point might be betting that the marginal cost of compute for Bitcoin will stop declining, and that the mining hardware cycle will become more like a traditional capital goods cycle—with longer replacement cycles and lower peak demand. That would hurt Lam’s growth premium, but it would also mean that miners with existing fleets (like Marathon, Riot) benefit from higher barriers to entry.
Furthermore, the export control regime is accelerating the “de-Americanization” of the global semiconductor supply chain. China’s third-phase Big Fund ($340B+) is pouring money into domestic equipment. While Lam’s technology advantage in high-end etch is safe for 5+ years, the medium-term risk is that Chinese foundries will learn to produce mid-tier mining ASICs (for 28nm and above) using domestic tools. That would erode Lam’s addressable market in the long run. Third Point’s exit is a signal that they see this risk as underpriced.
Takeaway: The Next Watch The next 6 months will tell us if Third Point is early or right. Watch Lam’s Q2 2025 order guidance—specifically the China revenue breakdown and the HBM tool mix. If orders decline sequentially, the miner hardware supply chain will face a 12-18 month lag before a new capex wave. Speed is the only moat when the gate opens. For now, the gate is closing.
Forensic accounting for the decentralized age requires reading between the lines of traditional finance filings. Third Point’s exit is a canary in the coal mine for crypto mining hardware costs. The question is whether the market will price in the structural shift before the next halving cycle begins.
