By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panel is divided on SanDisk's (SNDK) future, with concerns about memory cycle volatility, margin sustainability, and dependency on hyperscaler capex cycles. However, there's also optimism about AI-driven NAND demand and potential supply constraints due to 3D NAND yield issues.
Risk: Margin normalization and a slowdown in hyperscaler capex could compress SNDK's multiple well before 2027-28 tightness materializes.
Opportunity: Potential supply constraints due to 3D NAND yield issues could sustain high margins and drive enterprise-grade AI storage demand.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
The biggest winner in the stock market this year isn’t Nvidia (NVDA) or one of the Magnificent Seven stocks. And I’ll bet it won’t be SpaceX, Anthropic, or OpenAI, although those upcoming IPOs will get a lot of attention this year.
Instead, the biggest winner so far this year is SanDisk (SNDK), an AI infrastructure company on an outstanding two-year run. SNDK stock jumped nearly 560% in 2025 after the company completed its spinoff from Western Digital (WDC) in February. And so far this year, it's up another 650%.
Let me put it this way. Had you invested $10,000 in SNDK stock the day its split was complete just 16 months ago, you’d be sitting on $344,800 today. That’s an incredible return.
www.barchart.com
But amazingly, the run doesn’t appear to be coming to an end. SanDisk just got a bullish update from the financial institution Mizuho, which reiterated its “Outperform” rating on SNDK stock and raised its price target from $1,825 to $2,200.
“We continue to see AI as the driving force behind the supply-demand imbalance in the memory market, as we note increasing demand in (2027-2028) could add further pressure to the market,” analysts wrote.
Let’s take a closer look at SanDisk, which appears destined to be the best-performing stock in the S&P 500 this year.
SanDisk, which is headquartered in Milpitas, California, makes products for flash and advanced memory computer storage. The company was acquired by Western Digital in 2016, but in 2023, the companies announced that SanDisk would be spun back off. The newly emerged company includes SanDisk's and Western Digital's flash products with an emphasis on solid-state drives (SSDs), memory cards, and USB drives.
SanDisk’s storage technology can be used in phones, laptops, cameras, and game systems. The company’s biggest revenue stream is its edge computing division, which supplies storage for cars, drones, security cameras, and other smart devices that create data. But the greatest growth opportunity comes from its data center division, and that’s what’s drawing the most investor interest right now. Shares in the last 12 months are up over 4,300%, and it is by far the best-performing stock in the S&P 500 ($SPX) this year.
Story Continues
| | | | --- | --- | | S&P 500 Stocks | Year-to-Date Return | | SanDisk | 650% | | Micron Technology (MU) | 225% | | Dell Technologies (DELL) | 203% | | Seagate Technology (STX) | 207% | | Intel (INTC) | 223% |
Despite that run, SNDK isn’t tremendously expensive. Its forward price-to-earnings ratio is 26.3—high for the company, but fellow storage stocks Seagate and Western Digital have forward P/Es greater than 50. And the forward P/E of the S&P 500 is 22, meaning you are paying only a small premium for SNDK stock compared to the greater index.
Nothing would put a damper on a stock’s momentum like an earnings miss. But fortunately for investors, SanDisk has been a consistent winner when it steps to the earnings podium.
Revenue for the fiscal third quarter of 2026 (ending April 3, 2026) was $5.95 billion, up 251% from a year ago. The company posted net income of $3.61 billion, an increase of 287% from last year, and earnings of $23.41 per share versus expectations of $14.66 per share.
The quarter saw explosive growth in SanDisk’s data center division, which grew its revenue by 233% on a sequential basis and 645% from a year ago.
| | | | | | | | --- | --- | --- | --- | --- | --- | | Revenue | Q3 2026 | Q2 2026 | Q/Q Growth | Q3 2025 | Y/Y Growth | | Data Center | $1.467 billion | $440 million | 233% | $197 million | 645% | | Edge | $3.663 billion | $1.678 billion | 118% | $927 million | 295% | | Consumer | $820 million | $907 million | (10%) | $571 million | 44% | | Total Revenue | $5.950 billion | $3.025 billion | 97% | $1.695 billion | 251% |
“This quarter marks a fundamental inflection point for SanDisk—where our technology leadership is enabling a deliberate shift in our mix toward the highest-value end markets, led by data center,” CEO David Goeckeler said. “We are also advancing to a new business model built on multi-year customer engagements backed by firm financial commitments. Together, this transformation is driving structurally higher and more durable earnings power.”
SanDisk issued fourth-quarter guidance calling for revenue in the range of $7.75 billion and $8.25 billion, with gross margins between 79% and 81%. The company is projecting earnings per share of $30 to $33.
Simply put, analysts are expecting more of the same from SanDisk. Of 22 analysts who cover the stock, 19 of them have “Buy” ratings and the other three recommend holding. The consensus price target of $1,863.06 is lower than Mizuho’s bullish outlook but still represents potential upside of 8%. If you go by the Mizuho target of $2,200, then you’re looking at a potential upside of 28%. And the most bullish target of $3,250 implies dramatic gains of 89%.
At this point, SanDisk has risen so quickly—and has so much projected staying power—that it’s an ideal candidate for a stock split soon to make shares more accessible to retail investors. But in the meantime, don’t hesitate to use fractional shares to scoop up shares of SanDisk stock now.
www.barchart.com
On the date of publication, Patrick Sanders had a position in: NVDA. All information and data in this article is solely for informational purposes. This article was originally published on <a href="https://www.barchart.com/story/news/2423975/this-ai-infrastructure-stock-is-up-650-this-year-wall-street-still-sees-more-upside?utm_source=yahoo&utm_medium=syndication&utm_content=footer_link">Barchart.com</a>
Four leading AI models discuss this article
"SanDisk's valuation already embeds flawless execution through the next memory upcycle, leaving little margin for the supply response that has ended every prior NAND boom."
SanDisk's reported 645% YoY data center revenue jump and 79-81% gross margin guidance signal AI-driven NAND demand, yet the article's timeline of a 2023 spinoff from WDC and SNDK's independent S&P 500 status contradicts known corporate history. Forward P/E of 26.3x looks cheap versus Seagate's 50x+, but memory cycles have repeatedly crushed even high-growth names when supply catches demand. Mizuho's $2,200 target assumes 2027-2028 tightness materializes without execution slippage or inventory digestion. The 4,300% twelve-month run already prices in near-perfect outcomes.
Persistent AI capex from hyperscalers could extend the imbalance well beyond 2028, allowing SanDisk to sustain 50%+ margins and justify further multiple expansion despite historical sector mean-reversion.
"Sustained AI-driven data-center demand is the key make-or-break factor; without it, SanDisk's current valuation risks meaningful downside."
The article treats a spin-off rally as a durable AI infrastructure bet. SanDisk benefits from data-center storage demand, but it glosses over memory-cycle volatility and the risk that AI capex normalizes or memory pricing worsens. With forward P/E around the mid-20s and a huge YTD move, any slowdown in cloud/storage budgets or quicker price relief could trigger multiple compression. The 2027–28 demand thesis is plausible but not guaranteed; a miss or slower data-center growth could undermine the upside even if AI remains a tailwind. The optimism may also hinge on optimistic spin-off benefits requiring execution that isn’t guaranteed.
The strongest countercase is that this rally is largely multiple-expansion on a hot AI story rather than a proven, durable earnings trajectory; any shift in memory pricing or cloud capex could snap the stock back.
"SanDisk's current gross margin guidance is likely unsustainable and reflects a cyclical peak rather than a permanent structural shift in the memory industry."
The explosive growth in SanDisk's (SNDK) data center revenue—up 645% year-over-year—is undeniably impressive, but investors must look past the headline growth. A 79-81% gross margin guidance is historically anomalous for a memory/storage manufacturer, suggesting either a temporary pricing power anomaly due to extreme supply constraints or a fundamental misclassification of the business model. While the forward P/E of 26.3 looks attractive compared to peers like Micron (MU), it assumes these hyper-growth margins are sustainable. Memory is notoriously cyclical; if this is a 'super-cycle' peak, the valuation will compress violently once supply catches up to demand.
The bull case relies on a permanent structural shift in memory pricing, but if history holds, the current supply-demand imbalance will inevitably trigger massive capacity expansion, leading to a classic commodity-style margin collapse.
"SNDK's fundamentals are strong, but the stock has already capitalized most of the AI infrastructure thesis; further upside depends on whether hyperscaler capex accelerates into 2027-2028 or begins to moderate, a binary outcome the article treats as settled."
SNDK's 4,300% return in 12 months is extraordinary, but the article conflates two separate phenomena: a spinoff revaluation (Feb 2024) plus genuine data center demand. The Q3 2026 earnings are real—645% YoY data center growth, 59% net margin—but the forward P/E of 26.3x assumes this hypergrowth sustains. The guidance for Q4 (77-81% gross margins) is unusually wide, suggesting uncertainty. Mizuho's $2,200 target implies ~12% annualized upside from here; the stock has already priced in most near-term AI tailwinds. The real risk: SNDK's growth is contingent on sustained capex cycles from hyperscalers, which are cyclical and subject to demand destruction if AI ROI disappoints.
A 26.3x forward P/E on a stock that has already run 4,300% in 12 months leaves minimal margin of safety; even a single quarter of deceleration in data center growth (which is plausible as hyperscalers optimize inventory) could trigger a 30-40% drawdown, and the article provides no valuation context for what happens if growth normalizes to 50-100% YoY instead of 600%.
"AI NAND premiums offer no lasting buffer against capex pauses that reset multiples before the projected supply tightness arrives."
Gemini's dismissal of 79-81% margins as anomalous ignores that AI-optimized NAND commands structural premiums absent in prior cycles. Yet linking this to Claude's hyperscaler capex dependency, even modest 2025 budget normalization could collapse the 645% growth base, forcing the 26.3x forward P/E toward historical memory troughs of 12x before Mizuho's 2027 tightness materializes.
"Margins are unlikely to stay at 79–81%; a reversion is likely as supply responds, risking multiple compression before long-run AI-demand cycles materialize."
Gemini rightfully questions sustainability of a 79–81% gross margin, but the deeper flaw is assuming AI tailwinds alone keep pricing tight. Even if data-center demand stays hot, memory margins have historically retracted as supply responds. The missing risk: hyperscalers digestion of inventory and a 2025–26 capex pause could push gross margins back toward 50–60% and compress SNDK’s multiple well before 2027–28 tightness materializes.
"Technical limitations in NAND scaling create a structural supply floor that justifies sustained high margins regardless of cyclical demand."
Gemini and ChatGPT focus on margin normalization, but both ignore the physical reality of NAND scaling. We are hitting a wall in 3D NAND layer counts where yield loss on high-density chips creates a permanent supply floor. This isn't just a cyclical demand spike; it's a technical supply bottleneck. If yields remain sub-optimal at 300+ layers, those 80% margins aren't anomalies—they are the new cost of entry for enterprise-grade AI storage.
"A yield wall is only a durable margin support if we're already hitting it; if SNDK has headroom, this becomes a cyclical constraint, not structural."
Gemini's 3D NAND yield-wall argument is testable but needs specifics: what's the actual layer count ceiling, and does SNDK's current fab capacity already reflect this constraint? If yields are already sub-optimal at 300+ layers, margins should already be depressed—yet they're expanding. This suggests either yields are better than claimed, or SNDK isn't maxed out yet. If the latter, Gemini's 'permanent supply floor' collapses once yields improve or new capacity comes online. The physics argument only holds if we're already at the wall.
The panel is divided on SanDisk's (SNDK) future, with concerns about memory cycle volatility, margin sustainability, and dependency on hyperscaler capex cycles. However, there's also optimism about AI-driven NAND demand and potential supply constraints due to 3D NAND yield issues.
Potential supply constraints due to 3D NAND yield issues could sustain high margins and drive enterprise-grade AI storage demand.
Margin normalization and a slowdown in hyperscaler capex could compress SNDK's multiple well before 2027-28 tightness materializes.