US memory giant Micron briefly surpassed Meta and Tesla in intraday market cap on June 26, becoming Wall Street's hottest AI infrastructure concept stock; although it fell back to roughly match the two giants on the 27th, the "Next Nvidia" label has spread widely. Citigroup analyst Christopher Danely has raised Micron's target price to $200.
Citigroup Analyst Danely Raises Micron Target Price to $200
Citigroup analyst Danely has raised Micron's target price to $200, noting that "AI demand has created an unprecedented memory super cycle." Evercore ISI analyst Lipacis pointed out that Micron's strategic partnership agreement with Anthropic represents an unprecedented model in semiconductor industry history: AI model developers directly binding long-term supply contracts with memory manufacturers, a vertical integration that changes the memory makers' former passive position of simply following the economic cycle.
Micron's CEO stated in the most recent earnings call: "We are fundamentally transforming our business model," shifting from a commodity supplier subject to the cyclical economy to a strategic partner locking in revenue with long-term supply contracts.
IDC Points Out HBM Shortage to Persist Until 2027
High Bandwidth Memory (HBM) is a core component of Nvidia GPUs; each H200/B200 GPU requires 6 to 8 HBM stacks, and a single AI server can consume hundreds of GB of high-bandwidth memory. As cloud providers such as AWS, Google Cloud, and Microsoft Azure simultaneously expand computing capacity at scale, global HBM production capacity is severely undersupplied.
IDC notes that this shortage will persist until 2027 and has spilled over into consumer electronics: Apple was recently forced to raise prices on MacBooks and iPads, precisely due to surging memory costs.
Historical records in the memory industry are worth noting: over the past 30 years, the DRAM market has experienced at least five major boom-and-bust cycles; Samsung and SK Hynix are also expanding HBM production, and historically, every time supply caught up with demand, profit erosion occurred.
Micron Has Established HBM Packaging Hubs in Houli, Taichung and Guishan, Taoyuan
Micron has a massive investment footprint in Taiwan. Its A3 plant in Houli, Taichung and the Guishan plant in Taoyuan are both advanced DRAM and HBM packaging hubs; last year, Micron heavily procured Taiwanese equipment and materials, benefiting the memory packaging, testing, substrate, and test interface supply chains. SK Hynix recently reportedly plans to raise about $29.4 billion through a US IPO to expand HBM production, indicating that the AI memory race has extended from Asia to global capital markets.
Frequently Asked Questions
Why has HBM become the core bottleneck in the AI market?
High Bandwidth Memory (HBM) is an essential core component of Nvidia GPUs; each H200/B200 GPU requires 6 to 8 HBM stacks, forming the necessary foundation for AI training and inference computing power. The simultaneous large-scale expansion of computing capacity by global cloud providers has made HBM supply unable to meet demand, with IDC noting the shortage will persist until 2027.
What are the main assessments of Micron by Citigroup and Evercore ISI?
Citigroup analyst Danely has raised Micron's target price to $200, citing an unprecedented AI demand super cycle. Evercore ISI analyst Lipacis noted that Micron's strategic partnership agreement with Anthropic represents a new vertical integration model where AI model developers directly bind long-term supply contracts with memory manufacturers.
What identified historical risks does Micron's HBM growth face?
The memory industry has experienced at least five major boom-and-bust cycles in the past 30 years; Samsung and SK Hynix are also expanding HBM production, and historically, profit erosion occurred every time supply caught up with demand. Micron's CEO has stated that the company is transforming into a strategic partner through long-term supply contracts to reduce cyclical risk.
