Asymco’s Horace Dediu argues Apple’s scale, long-term supplier leverage, and cash reserves insulate it from the 2026 memory price spike hitting marginal players hardest.
Key Takeaways
Memory is moving from 15% to 40% of device BOM in 2026; the spike is concentrated in spot/variable pricing, not long-term base-load contracts Apple negotiates.
Apple’s multi-year lead-time contracts and volume commitments give it leverage to threaten supplier exclusion post-cycle, since memory booms always precede busts.
Samsung is reportedly earning more from memory now than Nvidia earns from GPUs, making DRAM the most valuable semiconductor product of 2026.
Apple could absorb margin compression (49% to 45% gross margin) while locking up supply, forcing smaller Android OEMs losing money at $800/unit to exit the market.
A speculative “Neo” low-end iPhone at $499 is floated as a potential low-end disruption play once the memory market normalizes and competitors are weakened.
Hacker News Comment Review
Commenters split on Dediu’s framing: skeptics called it “Apple fanfiction,” questioning whether cyclical bust logic actually constrains suppliers flush with AI-driven demand.
A practical counter-strategy raised repeatedly is vertical integration: Apple buying Micron outright or funding dedicated production lines at existing fabs to eliminate third-party leverage entirely.
Software efficiency was cited as the most immediate lever, with one ex-Apple kernel engineer noting Apple has repeatedly compressed RAM requirements across OS generations without user-visible regression.
Notable Comments
@nikhizzle: Ex-Apple kernel engineer; says Apple’s primary response will be OS-level memory efficiency improvements, a pattern repeated across multiple product generations.
@thisislife2: Argues soldered RAM is the structural root cause; unsoldered upgradeable memory would shift cost burden to customers and improve repairability.
