SanDisk benefits as agentic AI explodes context data. Most of that KV cache cannot live in HBM, so it spills to QLC NAND, which is slower but the only tier economical enough at that scale. x.com
Trade Whisperer
Memory Engineer & Data Scientist, Founder of Blue Pink Candles
Sandisk's business model is shifting structurally toward durable earnings, which points to a re-rating. CFO Luis Visoso said at the 2026 Investor Day that Sandisk has signed multi-year agreements with eight customers covering about two-thirds of FY2028 bits, supporting more predictable revenue, better cash flow visibility and durable earnings growth. Sandisk expects to return 100% of excess cash to shareholders after investing in the business. x.com
Agentic AI's repeated context favors cache memory and storage demand for Micron, SK hynix and Sandisk over fresh GPU computation. Per @demian_ai, about 95% of input tokens for DeepSeek V4.1 Flash on OpenRouter were cache hits; storage matters unless routing and caching keep almost everything in GPU memory. New products such as 3D DRAM and HBM-on-logic could be the final trigger for a full memory-stock re-rating. x.com x.com
SanDisk is attractive purely as a QLC NAND play in data centers, even though its high-bandwidth flash (HBF) remains unappealing. x.com
Sandisk's growth is coming entirely from higher prices, not more bits shipped, which makes it hard to sustain. Revenue grew 97% quarter over quarter as almost every hyperscaler raised capex to cover rising prices. Its roughly 80% gross margin is excessive and customers resent it; as long as demand far exceeds supply it can keep both, but supply eventually catches up and margins will erode. The open questions are how long new fabs take and whether Sandisk or rivals add the supply. x.com x.com … Show more
SanDisk's NAND story carries AI demand its own deck does not price. Of the 1.2 zettabytes on SanDisk's AI slide, 420 exabytes is KV cache, a write stream the deck leaves unpriced. Work with Rick Xie sorts out which AI bytes actually belong on flash. x.com
Kioxia and SanDisk are the purer bets on NAND and enterprise SSDs, with more torque to QLC, KV-cache and offload storage, and Kioxia trades at a far lower multiple. Enterprise SSD is where the memory supercycle is being priced, as US cloud providers raised eSSD demand forecasts and QLC emerges as the cost-efficient growth driver, per a post citing TrendForce. x.com
SanDisk's investor-day projections look too rich, as forecasting four years out on flash offload in AI inference is speculative. Measured against an 18TB attach-rate max bull case, now more like 8-10TB, they imply SanDisk assumes KV cache grows with users and context, many HBF customers, which looks unlikely now, or rock-solid four-year long-term agreements. The core disagreement is how big and how fast flash offload becomes ubiquitous. x.com x.com
SanDisk CEO David Goeckeler's report of customer signals beyond fiscal 2030 adds storage to the long-duration AI hardware-demand case. He says those signals come through new-business-model partnerships; Micron and SK hynix in memory and Lumentum in photonics were already cited as signs of demand beyond 2030. x.com
Sandisk benefits because AI needs more storage, and long-term customer commitments could make NAND demand and pricing more predictable. x.com
SanDisk's 2030 guidance has zero credibility, like Salesforce's or ServiceNow's long-range targets, and looks like its Oracle-RPO moment. A potential compute overbuild could arrive well before then, along with massive new memory capacity from everyone, including China. The guidance is still good for a big one-day gain. x.com x.com
SanDisk could benefit as DeepSeek's new memory architecture significantly increases the importance of NAND flash storage and SSDs in AI inference. x.com