BN
TechAI Desk•3 views

Wall Street Bulls Boost Micron Confidence: AI Drives Memory Demand

Major Wall Street banks are expressing strong confidence in Micron Technology, arguing that the market underestimates the long-term memory chip demand fueled by AI development. Analysts highlight a 'virtuous cycle' where larger AI models necessitate more memory, particularly in the key-value cache. A critical concern noted is the constrained supply of DRAM, which is limited to a few major manufacturers. Furthermore, reports indicate that the overall data center sector is expected to see massive growth, with high absorption rates anticipated even as capacity doubles through 2030. Several banks have issued high price targets, reflecting their bullish outlook on memory sector pricing power.

Ad slot
Wall Street Bulls Boost Micron Confidence: AI Drives Memory Demand

Major Wall Street investment banks are expressing heightened optimism regarding Micron Technology, suggesting the market is underestimating the sustained, long-term demand for memory chips driven by the Artificial Intelligence (AI) boom.

Key Analyst Predictions and Price Targets

Several leading financial institutions have significantly raised their outlook for Micron. Key highlights include:

  • DA Davidson: Doubled down on a $1,000 price target for Micron, citing increased conviction in the stock's buy rating.
  • Deutsche Bank: Also set a $1,000 price target, attributing the confidence to favorable cyclical dynamics and stable through-cycle returns.

The AI 'Virtuous Cycle' Driving Memory Needs

Analysts point to a self-reinforcing cycle within AI development that heavily favors memory components. Gil Luria of DA Davidson detailed this 'virtuous cycle' for memory manufacturers:

Ad slot
  • Model Scaling: Larger AI models inherently require more memory.
  • Key-Value Cache: The demand for memory in the 'key-value cache'—an intermediate step in Large Language Model (LLM) processing—is expected to exceed current market estimates.
  • Context Length: Longer context windows lead to better model intelligence, which in turn enables even larger and more complex models, perpetuating the demand cycle.

DRAM as the Critical Bottleneck

A significant theme across reports is the increasing importance and constrained supply of memory chips, particularly DRAM.

  • Shifting Pressure Point: Memory chips are becoming the most sensitive pressure point in the current AI buildout, moving focus away from CPUs.
  • Limited Supply: DRAM output is reportedly limited to three key players: Samsung, SK Hynix, and Micron, each with constrained production capacity.
  • Pricing Power: Mizuho estimates that pricing power for DRAM and NAND memory could climb substantially, projecting potential year-over-year increases of around 35% to 51% for these segments, driven by strong AI demand and tight supply.

Data Center Growth Outlook

Beyond memory specifics, the broader infrastructure supporting AI remains robust:

  • Capacity Growth: Barclays projects that data center capacity is poised to double between 2025 and 2030.
  • Absorption Rate: Despite this doubling capacity, analysts anticipate that the majority of new capacity will be absorbed, likely through pre-leasing.
  • Deployment Trend: The growth is expected to manifest not just in massive central hubs, but also in more traditional, geographically distributed colocation deployments as enterprise AI accelerates.
Ad slot