As of July 9 Eastern Time, Micron Technology shares rose 4.42% to $991.64, signaling a potential end to the recent correction. This surge follows the company's commitment to invest $250 billion in US-based memory manufacturing and R&D by 2035, reflecting strong confidence in AI-driven demand.
While fiscal 2026 results show exponential revenue and EPS growth, the massive capital expenditure poses long-term risks regarding valuation and supply overcapacity. Technically, support at $900 remains critical; a sustained break above $1,050 resistance is necessary to target previous highs and potentially reach $1,400.
Sandisk (SNDK) has pulled back by 20% from recent highs, although in this case, 20% can be considered peanuts given the shares are still up by 683% year-to-date.
Nevertheless, market watchers are starting to wonder whether the pullback is just the first phase of a wider retreat for high-flying memory stocks that have soared to incredible heights, or whether it is merely a temporary blip before momentum builds again.
For those mulling this dilemma, an investor with the pseudonym of Summit Research (SR) has a solution. “The latest stock pullback could represent the last discounted opportunity to participate in SNDK’s final re-rating for AI NAND demand upside before the broader industry supply-demand imbalance eases in 2028,” SR said.
The investor believes Sandisk remains a “high-conviction beneficiary of the emerging AI-driven NAND supercycle.”
SR’s core investment thesis is that AI infrastructure is expanding beyond traditional compute-focused architectures. As AI models grow more complex and require larger context windows and greater KV cache capacity, storage is becoming a more important part of the “AI memory hierarchy.” While DRAM and high-bandwidth memory (HBM) remain critical for immediate data processing, NAND flash is moving closer to compute by supporting higher-performance warm storage applications.
The investor believes Sandisk is well positioned to capitalize on this shift through its expanding data center NAND portfolio. The company’s BiCS8-based QLC enterprise SSDs are designed to address growing demand for high-density storage in AI data centers, while its latest BiCS10 TLC 3D NAND strengthens its position in more performance-sensitive applications. BiCS10 delivers improvements in density, bandwidth, latency, and power efficiency, making it more relevant for AI inference workloads where performance and energy efficiency are becoming key constraints.
Comments