F22

$MU I worked 21 years as an HBM, DRAM & NAND engineer. AMA is open. Ask me anything. I'll drop rare insights where I can.

There are people still doubting if $META can effectively use AI for its core business. Do they think Meta just magically snapped its fingers or accelerated from 16% to 32% growth by chance? Just listen to an earnings call. AI is dramatically helping their core. Right now.

Micron $Mu is currently undergoing a structural transformation to reduce its historic dependence on the volatile consumer PC and smartphone cycles.  Micron has moved aggressively to position itself as an "AI powerhouse," shifting its center of gravity toward Enterprise and AI Data Centers.  The Strategic Pivot away from Consumer Historically, Micron's revenue was tethered to the "boom-bust" cycles of the consumer sector. To break this, the company has taken several unprecedented steps:  1. Exiting Consumer Brands: In late 2025, Micron announced it would exit the Crucial consumer business (retail RAM and SSDs) by February 2026. This move was designed to reallocate manufacturing capacity toward higher-margin enterprise products.  2. Wafer Displacement: High Bandwidth Memory (HBM) for AI servers requires approximately three times the wafer capacity of standard DRAM. By focusing on HBM, Micron naturally reduces the oversupply of "commodity" consumer DRAM, which helps stabilize prices across the board.  3. Enterprise Mix: Data center bit demand is projected to exceed 50% of the industry total for the first time in 2026, with enterprise SSDs now making up over 60% of Micron's NAND revenue.  4. Seeking "Durable" Revenues Micron is attempting to replace "spot price" volatility with Strategic Customer Agreements (SCAs). These are multi-year, fixed-price contracts with major hyperscalers (like NVIDIA, Microsoft, and Amazon).  5. The Goal: By locking in these agreements, Micron aims to create a "margin floor." Instead of being repriced every quarter by the open market, they operate more like a long-cycle industrial business. 6. High Bandwidth Memory (HBM): Products like HBM3E and the newly ramped HBM4 command significantly higher Average Selling Prices (ASPs) and margins (projected to hit 81% in late 2026) compared to legacy consumer chips.  Is the "Cyclical Nature" Truly Gone? While the pivot is real, experts remain split on whether the cycle is truly "broken" or just "extended": The Bull Case: The "Memory Wall" in AI is a structural bottleneck. AI models need more memory just as much as they need faster processors, creating a permanent demand floor.  The Bear Case: Even in the enterprise sector, cycles exist. If big tech companies (the "Magnificent Seven") eventually slow their data center build-outs in 2027 or 2028, the industry could still face a "digestive period" or oversupply. Takeaway Micron is using the AI transition to fundamentally change its business model from a "commodity chip maker" to a "strategic infrastructure partner," aiming for a more predictable and premium revenue stream.

Micron $Mu is currently undergoing a structural transformation to reduce its historic dependence on the volatile consumer PC and smartphone cycles.  Micron has moved aggressively to position itself as an "AI powerhouse," shifting its center of gravity toward Enterprise and AI Data Centers.  The Strategic Pivot away from Consumer Historically, Micron's revenue was tethered to the "boom-bust" cycles of the consumer sector. To break this, the company has taken several unprecedented steps:  1. Exiting Consumer Brands: In late 2025, Micron announced it would exit the Crucial consumer business (retail RAM and SSDs) by February 2026. This move was designed to reallocate manufacturing capacity toward higher-margin enterprise products.  2. Wafer Displacement: High Bandwidth Memory (HBM) for AI servers requires approximately three times the wafer capacity of standard DRAM. By focusing on HBM, Micron naturally reduces the oversupply of "commodity" consumer DRAM, which helps stabilize prices across the board.  3. Enterprise Mix: Data center bit demand is projected to exceed 50% of the industry total for the first time in 2026, with enterprise SSDs now making up over 60% of Micron's NAND revenue.  4. Seeking "Durable" Revenues Micron is attempting to replace "spot price" volatility with Strategic Customer Agreements (SCAs). These are multi-year, fixed-price contracts with major hyperscalers (like NVIDIA, Microsoft, and Amazon).  5. The Goal: By locking in these agreements, Micron aims to create a "margin floor." Instead of being repriced every quarter by the open market, they operate more like a long-cycle industrial business. 6. High Bandwidth Memory (HBM): Products like HBM3E and the newly ramped HBM4 command significantly higher Average Selling Prices (ASPs) and margins (projected to hit 81% in late 2026) compared to legacy consumer chips.  Is the "Cyclical Nature" Truly Gone? While the pivot is real, experts remain split on whether the cycle is truly "broken" or just "extended": The Bull Case: The "Memory Wall" in AI is a structural bottleneck. AI models need more memory just as much as they need faster processors, creating a permanent demand floor.  The Bear Case: Even in the enterprise sector, cycles exist. If big tech companies (the "Magnificent Seven") eventually slow their data center build-outs in 2027 or 2028, the industry could still face a "digestive period" or oversupply. Takeaway Micron is using the AI transition to fundamentally change its business model from a "commodity chip maker" to a "strategic infrastructure partner," aiming for a more predictable and premium revenue stream.

RECORD NUMBER OF AMERICANS LEAVING US FOR BETTER QUALITY OF LIFE ABROAD: WSJ

Buffett’s portfolio, visualized. Top 5 holdings drive the majority of Berkshire’s equity exposure with $AAPL leading, alongside long-time pillars $AXP and $KO. Concentration, patience, and conviction over decades.

Warsh will be confirmed in 2 weeks. May the printing begin...