Memory Chips

Let me share a few thoughts from my mental ledger. Let me start with my assumptions. First, this round of "tide" in memory is still riding the tail of the rising swell—NAND and DRAM contract prices remain in the historical high band in Q2, and pull-through from AI servers and enterprise SSDs hasn't broken stride. I'm anchoring Q2 revenue around the 8.15 billion mark, with sequential growth of nearly 20%, fueled by the ramp of automotive-grade memory stacked on top of continued long-term orders from key accounts. Second, on gross margin—I'm no longer betting it keeps pushing higher. Around 51% feels like a healthier level to me. Q1's 53.3% already captured the sweetest slice of the price elasticity; for Q2, assuming prices hold flat at the high end and product mix continues tilting toward enterprise and automotive, GM can stabilize at 51%—but that's the ceiling. Then for Q3, I'm modeling 8.62 billion in revenue and a 48.5% gross margin. Revenue is still climbing, but the growth rate is visibly hitting the brakes; gross margin takes a step down. This isn't me being bearish—it's a measure of respect I have for the cycle. Historically, memory has never strung together four consecutive quarters of expanding gross margins; once there's any loosening signal on the supply side, the price side will be the first to react. I'd rather leave a trace of the "ebbing tide" in the numbers now than wait for the Q3 report to drop and have to scramble for backfills. Net profit across these two quarters lands at 3.56 billion and 3.42 billion respectively—sequentially essentially flat, or even slightly down. That's consistent with my call on gross margin topping out, and it leaves operating expense room for management to keep funding R&D and expanding automotive-grade production lines. The biggest risk is really just one thing: if upstream suppliers' capex materializes in a concentrated wave in 2H this year, and channel inventory starts to build on top of that, this cycle could turn before Q4. Sure, automotive-grade and enterprise long-term agreements can dampen the volatility to some extent—but a tide is a tide, and we surfers have to keep our knees bent.