Big Tech CEOs Are All Sounding the Same Alarm — And It Points Directly to Micron and Sandisk
The Boardroom Consensus No One Expected
When the leaders of four of the world's most powerful technology companies independently highlight the same supply constraint during their earnings calls, investors should pay attention. Across Apple (AAPL), Tesla (TSLA), Amazon (AMZN), and Nvidia (NVDA), a single theme dominated the conversation this earnings season: memory. Not artificial intelligence models, not energy infrastructure, not even GPUs — but high-bandwidth memory (HBM), DRAM, and NAND flash storage.
What was once considered a mundane commodity sector has quietly become one of the most critical bottlenecks in the global AI build-out, and the executives managing trillion-dollar companies are now talking about memory chips the way traders once talked about oil.
What Each CEO Actually Said
Apple's outgoing CEO Tim Cook set the tone by telling investors that the company paid more for memory in its March quarter than in the prior December quarter — and expected to pay "significantly more" in the June quarter, with even higher costs anticipated in September. Cook described the current pricing environment as "a 100-year flood." To offset the pressure, Apple has already raised prices on select Mac and iPad products, though Cook indicated that margin compression could deepen further in the current quarter.
Elon Musk took a more direct approach. During Tesla's earnings call, he specifically named Micron Technology (MU), thanking the chipmaker for providing Tesla with "a very significant allocation on reasonable terms given the pretty insane pricing of memory these days." Musk also attempted to frame the supply-demand dynamic in stark terms: memory production may be growing around 20% annually, but if demand is expanding closer to 200%, basic economics dictates that prices must continue rising.
At Amazon, CEO Andy Jassy revealed that the company had increased its capital expenditure budget for the year from roughly $200 billion to $220 billion — citing "the higher cost of memory" as a key driver. Even with that additional $20 billion, Jassy told investors that Amazon still won't be able to build enough compute capacity in the current year to satisfy existing demand, and that the shortage is expected to persist into 2027.
Jensen Huang's comments at Nvidia were perhaps the most striking in dollar terms. During Nvidia's fiscal 2027 second-quarter earnings call, Huang indicated that demand for the company's processors could support significantly more than 70% revenue growth in fiscal 2028 — but that current supply dynamics constrain the forecast to that level. To secure its pipeline, Nvidia has committed to $279 billion in supply and capacity arrangements through fiscal 2032, with those commitments described as "primarily related to the procurement of memory." Of that total, $267 billion is tied to supply contracts for the next two and a half years.
Why This Matters Beyond the Headlines
A common misreading of the memory price surge is that it signals the AI infrastructure cycle is overheating or nearing a peak. The evidence from these earnings calls suggests the opposite dynamic. Big tech players are not pulling back spending — they are accelerating it, even as memory prices climb sharply. If this were a fading trend, the rational response would be to defer capital expenditures and wait for prices to normalize. Instead, companies are locking in multi-year supply commitments and expanding their capex budgets.
Amazon is spending more because memory costs more and the company still needs additional rack capacity. Nvidia is guiding for record growth while simultaneously identifying memory supply as its primary constraint. Apple is absorbing higher input costs and passing some of that burden to consumers. Tesla and SpaceX are, in Musk's framing, simply grateful to secure allocations at all.
This pattern is consistent with what analysts describe as a durable supercycle — one in which demand so thoroughly outpaces supply that even elevated prices fail to cool purchasing activity.
The Knock-On Effect for Micron and Sandisk
The financial pain being absorbed by the tech sector's biggest spenders has a mirror image on the supply side. Micron Technology operates in a DRAM market with only three meaningful global suppliers, giving it significant pricing leverage. HBM — the highest-margin memory product category — is reportedly sold out well into next year. Meanwhile, Sandisk (SNDK) specializes in NAND flash storage, a segment that is seeing prices pulled upward as data center buildouts create downstream demand for storage capacity.
The impact of rising memory costs is already showing up in the financials of the companies buying these chips. Apple has reported sequential gross margin compression. Nvidia has guided investors toward a gross margin profile in the low-70s percentage range, down from the mid-70s, as memory inflation works through its supply chain. Amazon and Tesla have seen cash flow figures weaken during the very quarters when their compute demand narratives were strongest.
Those costs are flowing directly into the revenue lines of memory producers. Data indicates that both Micron and Sandisk are currently trading at forward price-to-earnings (P/E) multiples in the 6 to 7 range — historically modest valuations for companies operating in a segment that four major tech CEOs just described as critically undersupplied.
What to Watch Going Forward
The central question for investors tracking this space is how long the supply-demand imbalance persists. Musk's framing of ~20% supply growth against potentially 200% demand growth suggests the gap will not close quickly. Jassy's comment that Amazon's compute shortage extends into 2027 provides a timeline that spans multiple quarters. Nvidia's $267 billion in near-term memory procurement commitments signals that the company does not expect the constraint to ease anytime soon.
The next set of earnings calls and any updates to capital expenditure guidance from major hyperscalers will be critical indicators of whether this trajectory is holding. Supply announcements from memory manufacturers and any changes in HBM allocation timelines will also be closely watched by analysts covering the semiconductor sector.
Disclaimer: This article is for informational purposes only and does not constitute financial advice, investment recommendations, or an endorsement of any particular security or strategy. Always conduct your own research and consult with a qualified financial advisor before making investment decisions. Past performance is not indicative of future results.
Written by
Sarah ChenRelated Articles
AI Data Center Spending Could Hit $31.6 Trillion by 2050 — Four Companies Positioned Across the Supply Chain
Read more
NEWSU.S. Markets Slide as Middle East Tensions Flare, Oil Prices Surge — But August Closed Strong
Read more
NEWSGlobal Bond Yields Surge to Multi-Decade Highs as Middle East Conflict Reignites Inflation Fears
Read more