Data Centers Now Deliver a Third of Sandisk's Revenue -- $2.98 Billion in a Single Quarter

Sandisk’s fiscal Q4 datacenter revenue surged to $2.98B (about 1/3 of $8.97B total), up from $213M a year earlier, while the shift was reinforced by 10 multiyear “New Business Model” agreements with a minimum $93.9B in expected revenue at floor pricing and $16.5B of customer deposits. Profitability surged with gross margin expanding to 84.6% (from 26.2% a year ago) and quarterly net income swinging to $6.9B; full-year free cash flow improved from -$120M (FY2025) to $11.5B. Management guided FY1Q27 revenue to $10.3B–$10.8B (+15% to +20% sequentially) with gross margin of 83%–85%, though the stock is still down ~35% from its 52-week high and the market remains skeptical about sustainability of peak pricing.

Analysis

This is less a one-quarter earnings beat than an attempt to reframe NAND as a contracted infrastructure business. If the customer mix keeps shifting toward hyperscale/datacenter buyers, the market should lower the probability of the classic inventory bust and assign SNDK a higher trough multiple than legacy flash names. That also creates relative pressure on peers with more spot exposure: any competitor still leaning on consumer channels will look structurally lower quality in the next cycle.

The near-term risk is that investors confuse peak-margin optics with a permanent reset. Floors protect downside on contracted bits, but they do not immunize the remaining open-market volume; if spot NAND rolls over, earnings can still mean-revert quickly even with strong backlog. The key catalyst over 1-3 months is whether management can keep expanding contracted shipment coverage without sacrificing pricing discipline; the falsifier is a guide that implies coverage stalls or margins slip back toward the low-80s.

The contrarian view is that the market may be underestimating how much customer deposits improve financing visibility, which should support capex and partner commitments across the NAND supply chain. But that same visibility can also encourage capacity response, which is why this is not a clean scarcity story. Over 6-18 months, the bigger debate is not demand strength but whether SNDK has traded one cyclical driver for another: contractual rigidity that caps upside if the industry tightens again.

Key Decisions for Investors

• Long SNDK vs. short WDC as a 1-3 month pair trade: SNDK has the cleaner revenue-quality upgrade and should earn a premium if the market starts discounting fewer downside surprises. Risk: if NAND spot weakens broadly, the pair can converge against both legs; cover if SNDK loses post-earnings support or if WDC closes the valuation gap without SNDK follow-through.
• Buy SNDK on pullbacks rather than strength, with a 3-6 month horizon: the stock is likely to remain volatile as investors debate whether current margins are peak or durable. Reward/risk improves if the market gives back part of the post-print move while management reiterates contracted coverage and margin floors.
• Use a call-spread expression on SNDK instead of outright stock for 6-12 months: the upside is more likely multiple expansion than another explosive earnings step-up, while the downside is a rapid de-rating if pricing normalizes. Best if implied volatility stays elevated after earnings.
• Set an alert on next quarter's gross margin and contracted shipment coverage: if gross margin falls below the low-80% area or contracted coverage does not continue trending toward the stated FY27/FY28 targets, the thesis shifts from structural to cyclical.