The Wall Street Journal reports that memory chip manufacturers are reaping significant gains from AI infrastructure buildouts, with DRAM and HBM (High Bandwidth Memory) prices climbing as hyperscalers race to equip AI accelerators with high-speed memory. The dynamic reflects a fundamental shift in the memory cycle — historically brutal boom-bust — toward a tighter, AI-anchored demand floor.

The primary beneficiaries are pure-play memory names: Micron Technology (MU) and SK Hynix are the clearest direct plays, as both produce HBM used in Nvidia's H100 and B200 GPU stacks. Samsung is also in the mix but has lagged peers on HBM yield. QCOM appears in the enrichment data but is only tangentially linked — Qualcomm is a fabless logic chip designer, not a memory producer, and its $44.3B revenue base (+13.7% YoY) reflects smartphone and IoT chip sales rather than DRAM/HBM dynamics.

The bull case for memory names rests on a structural argument: AI capex spending from Microsoft, Google, Meta, and Amazon is creating a sustained HBM demand wave that is unlike prior consumer-driven memory cycles. HBM3e supply remains concentrated at SK Hynix and Micron, giving pricing power that was absent in the 2022–23 downturn.

The bear case is the memory industry's own history — capital cycles eventually catch up, Samsung is aggressively investing in HBM capacity, and any slowdown in hyperscaler capex (or Nvidia GPU shipment hiccups) could unwind the price surge quickly. Inventory builds at the customer level are a known risk. MU's next earnings print will be the key near-term catalyst to watch for forward pricing commentary.

QCOM's inclusion in this story's ticker enrichment is a data mismatch — its fundamentals (12.5% net margin, $5.01 diluted EPS) speak to a stable but unexciting logic chip business that does not directly move on memory price trends. The cleaner trade expression sits in MU, not QCOM.