The AI infrastructure frenzy has ignited the storage SSD to reach a super cycle trend! Will the surge of 300% in SanDisk (SNDK.US) within the year not yet come to an end?

As of Monday's closing in the US stock market, the global leader in SSD storage products, SanDisk (SNDK.US), continued its unprecedented strong upward trend since 2025, surging by nearly 12% at the close. Since the beginning of 2026, its stock price has skyrocketed by over 300%, far outperforming global stock market benchmarks. Throughout 2025, SanDisk's stock price soared an astonishing 580%, and this year, it continues to experience an epic surge akin to the "global storage chip myth," undoubtedly as the "super bull stock" of the global stock market, focusing on AI computing infrastructure. SanDisk's recent surge is not due to a single catalyst, but rather a resonance between the "passive buying brought about by index inclusion" and the "revaluation of fundamentals under the storage supercycle." On one hand, Nasdaq has announced that SanDisk will replace Atlassian on the Nasdaq 100 before the market opens on April 20, mechanically triggering the massive ETF and passive fund reallocation into the tracking index, naturally amplifying the stock price momentum in the short term. On the other hand, the more critical theme is Wall Street's continuous upward revision of SanDisk's target price for the next 12 months based on the hardcore logic of the so-called "storage supercycle." Evercore's initial coverage of SanDisk gave a target price of $1,200, a bullish scenario of $2,600, Citigroup raised it to $980, and Jefferies significantly raised the target price to $1,000, all betting on the explosive increase in enterprise-level DRAM/NAND storage demand driven by the craze in AI data center construction especially the tight supply and demand balance of SSDs and the profit reevaluation following the increase in the proportion of large AI data center businesses. DRAM and NAND soar together, storage chips become the hardest currency in the AI era According to the latest storage price survey by TrendForce Consulting, in the second quarter of 2026, due to DRAM manufacturers actively shifting capacity to HBM, server applications, and adopting a "price rise" strategy to close the price gap between various products, the overall contract price of conventional DRAM is estimated to still increase by 58%-63% month-on-month despite facing the risk of shipment reduction in the end market. The NAND Flash market continues to be dominated by AI and data center demand, with the effects of price increases across the product line persisting, and overall contract prices in the second quarter are expected to significantly increase by 70%-75% on top of the 100% increase in the first quarter. NAND Flash is the core storage medium for companies like Sandisk producing SSDs. Whether it is the massive TPU AI computing cluster led by Alphabet Inc. Class C or the massive NVIDIA Corporation AI GPU computing cluster, both rely on HBM storage systems integrated with AI chips, and in addition to HBM, these tech giants like Alphabet Inc. Class C and OpenAI must also purchase server-grade DDR5 memory and enterprise-level high-performance SSD/HDD storage solutions in large quantities for the rapid growth of AI data center business. Unlike Seagate and Western Digital Corporation focusing on dominating nearline high-capacity HDDs, SanDisk has long been focused on high-performance eSSDs. The three major storage chip manufacturers - Samsung Electronics, SK Hynix, and Micron are simultaneously dominant in multiple core storage areas: HBM, server DRAM (including DDR5/LPDDR5X), and high-end data center enterprise-level SSDs (eSSD). These storage chip leaders are the most direct beneficiaries in the "AI memory + storage stack" and are profiting from the "super dividend" of AI infrastructure led by tech giants such as Alphabet Inc. Class C, Microsoft Corporation, and Amazon.com, Inc. From a bottom-up hardware theory perspective, AI computing is fundamentally limited not only by computing power but also by "data transport capabilities." Whether it's NVIDIA Corporation's GPU or the TPU computing system, what truly determines the efficiency of large model training and inference is not just the number of Tensor Core/matrix units but also the bandwidth at which weights, KV caches, activation values, and intermediate tensors can be fed into the compute cores per second. Looking at the cross-analysis perspective of semiconductor and AI data center infrastructure, the reason why DRAM/NAND storage chips have placed themselves perfectly in the AI wave is that they are benefiting from both the expansion of training and inference. They are also a "universal toll station" that spans platforms, architectures, and ecosystems. When the AI era shifts from training-dominated to inference, agent, long context, and enhanced retrieval, the demand for capacity, bandwidth, power efficiency, and persistent data layers will only increase. AI is driving a surge in storage demand! Evercore's initial coverage of SanDisk gives an amazing target price Wall Street financial giant Evercore ISI started covering the storage giant, giving it an "outperform market" rating. What's more significant is that the Financial Institutions, Inc. gave an extremely high target price of $1,200, which is the highest on Wall Street, and in the bullish scenario of a global stock market, the target price is even more astonishing at $2,600. Based on SanDisk's closing price of $952.50 on April 13, the potential upside is approximately 26.0% for $1,200 and 173.0% for $2,600 in the next 12 months. The stock analysis team led by analyst Amit Daryanani at Evercore stated in an investor report, "We believe that SanDisk (SNDK) benefits from one of the most attractive areas in the AI computing infrastructure stackdata storage. The demand in this area is exponentially growing, while the supply is expected to remain extremely tight at least until calendar year 2028, if not longer. Although concerns about peak NAND prices and cyclicality persist in the market, we believe that this cycle is structurally tighter and more sustainable, supported by the near-endless demand driven by the AI data center construction frenzy and continued supply discipline, leading to strategic contract agreements (SCAs) that provide storage vendors with a higher price floor and prepayment cash (i.e., long-term strategic contract agreements between cloud computing companies and NAND/DRAM suppliers)." Despite the significant increase in SanDisk's stock price since the beginning of this year, Evercore believes that there is still further valuation and fundamental growth potential. The analyst team led by Daryanani stated, "The increasing exposure to enterprise/large-scale AI cloud computing business should provide a more favorable growth path and a more diversified business structure. Currently, large data center customers account for less than 15% of the total sales, but we expect this sector to accelerate its growth." Evercore believes that SanDisk should be able to maintain high profit margins in the long term and continue to improve its free cash flow. The firm also expects that SanDisk's management will return value to shareholders through stock buybacks. The analyst team led by Daryanani stated, "Although price remains the largest driver of changes in upside elasticity, we believe that cost execution and product restructuring are more sustainable contributing factors that will support its profitability at a structurally much higher level than during any previous growth cycle." SanDisk is expected to release its third-quarter financial results for the 2026 fiscal year on April 30, with the latest consensus expectations in the stock market being adjusted earnings of $14.23 per share on a revenue basis of $4.65 billion, and GAAP earnings per share of $13.62. The reason why SanDisk is seen as one of the biggest winners in this "storage supercycle" is first and foremost because it is highly leveraged and exposed to the upward cycle of NAND/enterprise-level SSD prices. TrendForce's survey shows that in the second quarter of 2026, the contract price of NAND Flash is expected to rise by an additional 70%-75% compared to the previous quarter, driven by strong demand from AI data centers, persistent supply shortages, and the shift in capacity towards higher-margin enterprise products; Wall Street's core prediction currently suggests that the mismatch between supply and demand for memory/flash may continue until around 2028, and the market still undervalues SanDisk's profit flexibility in this cycle. Therefore, SanDisk's stock price has surged since the beginning of 2026, and extreme revaluations have occurred over the past year - this is not just emotional hype but a reevaluation by the market of the pricing power in the "AI-driven storage." The second more fundamental reason is that SanDisk is not just benefiting from the "NAND price increase logic" but also from the truly scarce enterprise-level data storage layer in AI infrastructure. SanDisk's official promotional material clearly states that its enterprise SSDs are primarily aimed at AI data centers, commercial storage, AI/ML, HPC, and mixed workloads, with the SN861 NVMe SSD directly targeting AI/ML, HPC, and ultra-low latency high-throughput scenarios; the company's materials on "Empowering AI" for enterprise SSDs also emphasize that multiple key stages of the AI data cycle require high-performance data storage support. In other words, AI training and inference not only require AI GPUs/TPUs and HBM memory systems but also a large number of high-performance SSDs to handle data ingestion, caching, checkpoints, vector databases, logs, and inference workflows. This means that SanDisk is not just selling commodity NAND but higher value-added enterprise/data center SSDs. Evercore's benchmark target price of $1,200, defined SanDisk as one of the most attractive beneficiaries of "data storage" in the AI infrastructure stack, and believes that the proportion of large tech enterprise/cloud computing customers will continue to rise. In the past, the market was concerned that once NAND production demand peaked, companies like SanDisk and Kioxia would rapidly decline in profitability. However, this time, the consensus among top Wall Street research institutions such as Evercore, Bernstein, and Cantor is that there is stronger supply discipline, more long-term contracts, and cloud providers willing to prepay cash to secure supply, as well as an increase in the proportion of large enterprise/AI hyperscalers cloud business, making the profit structure more favorable. Evercore explicitly mentioned that SCAs are providing storage vendors with a price floor and prepayments; TrendForce also pointed out that AI demand is shifting the NAND industry from a "consumer electronics cycle" to an "enterprise-level SSD tight supply cycle," as AI data center demand diverts more storage capacity towards enterprise-level HBM and SSD, resulting in a continuous tight balance in the entire NAND market.