Key Takeaways
- Micron CEO Sanjay Mehrotra says the current memory chip shortage is only the opening act — 2027 and 2028 will be tighter than 2026.
- AI data centres are eating up DRAM and NAND capacity meant for phones, laptops and servers, via high-bandwidth memory (HBM) chips.
- Indian consumers are already seeing costlier RAM, SSDs and finished electronics, and the squeeze is expected to deepen, not ease.
- Micron’s own Sanand (Gujarat) assembly plant gives India a front-row seat to this global memory crunch, even though it doesn’t make raw chips there.
What Exactly Did Micron’s CEO Say About the Memory Chip Shortage?
Sanjay Mehrotra, the chief executive of US-based Micron Technology — one of only three companies left in the world that make advanced DRAM memory — told investors that the memory chip shortage currently squeezing the market is going to get worse, not better, as the decade rolls on. His exact framing: supply in 2027 and 2028 will be “much tighter” than it already is in 2026, a year that itself has seen some of the steepest memory price hikes in over a decade.
That’s a notable thing for a chipmaker’s boss to say out loud. Normally, executives talk up near-term demand and stay vague about years that far out, because nobody can really predict chip cycles with precision. Mehrotra going on record about 2027-28 tells you Micron’s order books, and its conversations with the handful of customers who buy memory in bulk — Nvidia, Dell, Samsung, Apple, the big cloud players — are already locked in well beyond the usual planning horizon.
Why Is There a Memory Chip Shortage in the First Place?
To understand this, you have to picture a memory fab the way you’d picture a multiplex. There are a fixed number of screens (wafer capacity), and once a film is booked in, that screen isn’t free for anything else for weeks. Right now, the blockbuster booking every available screen is artificial intelligence.
The AI Data Centre Effect
Training and running large AI models needs a special, expensive type of memory called High Bandwidth Memory, or HBM, which sits right next to AI processors like Nvidia’s GPUs. HBM uses far more silicon and far more manufacturing steps per gigabyte than ordinary DRAM. So when Micron, Samsung and SK hynix shift wafer capacity toward HBM to chase AI demand, there’s mechanically less room left on the line for the regular DRAM and NAND that goes into your laptop, phone or smart TV. Less of a shared resource, chasing unchanged or rising everyday demand, is the textbook recipe for a memory chip shortage.
DDR5 and the Legacy-Node Squeeze
Layered on top of that, the industry is mid-way through switching from the older DDR4 memory standard to DDR5. Factories can’t run both forever — at some point, older DDR4 lines get repurposed or shut, even while plenty of budget laptops and servers still need DDR4 chips. That transition always creates a temporary pinch. This time, it’s colliding head-on with the AI-driven HBM boom, which is why analysts tracking the memory chip shortage keep pushing back their forecasts for when prices will cool off.
How Long Will the Memory Chip Shortage Really Last?
Nobody has a crystal ball, but based on what Micron, SK hynix and Samsung have told investors over recent quarters, here’s a rough shape of where things stand and where they’re headed:
| Year | DRAM Supply Situation | NAND Supply Situation | Main Driver |
| 2026 | Tight, prices already climbing sharply | Tightening, though less severe than DRAM | AI/HBM demand plus DDR5 transition |
| 2027 | Tighter — Micron flags this as notably worse than 2026 | Tighter, new fab capacity still years from ramping | HBM demand growing faster than new fab investment |
| 2028 | Similarly tight or worse, per Micron’s guidance | Expected to remain constrained | Capacity additions announced today take 2-3 years to come online |
The key thing to notice in that table is the lag. Building a new memory fab, or even expanding an existing one, isn’t like adding a shift at a garment factory — it takes two to three years and billions of dollars before a single new wafer comes out the other end. That’s exactly why Micron’s CEO feels comfortable calling the memory chip shortage for 2027 and 2028 already: the fabs that would fix it haven’t broken ground yet.
What’s Micron’s Role, and Why Mention India at All?
Micron is the only major DRAM and NAND maker headquartered in the United States; the other two serious players, Samsung and SK hynix, are South Korean. For readers following India’s chip ambitions, Micron is also the company building a semiconductor assembly, test, marking and packaging (ATMP) unit at Sanand, Gujarat, backed by incentives under India’s Semiconductor Mission. That plant doesn’t fabricate raw memory wafers — the actual silicon still comes from Micron’s fabs in the US, Japan, Taiwan and Singapore — but it packages and tests finished chips, and it’s the clearest sign yet that India is being wired into the global memory supply chain, not just the demand side of it. You can read more about the company’s global footprint on Micron’s company page on Wikipedia, which tracks its fabs and expansion history.
That matters for the India angle on this story: a tighter global memory chip shortage through 2027-28 doesn’t just hit Micron’s balance sheet. It flows straight into what Indian consumers and businesses pay for laptops, servers, smartphones and even cars, since modern vehicles now carry significant memory content for infotainment and driver-assist systems.
How Does This Memory Chip Shortage Affect Indian Consumers and Businesses?
India imports almost all of its finished memory chips and most of its DRAM-heavy devices. When global DRAM and NAND prices rise, the effect shows up here with a short lag — usually one to two quarters — in:
- Laptop and desktop PC prices, especially mid-range configurations with 16GB or 32GB RAM.
- Smartphone prices, particularly mid-range and premium models with higher RAM/storage variants.
- Data centre and cloud hosting costs, since Indian IT and SaaS companies rent server capacity that’s priced partly on memory costs.
- Consumer electronics like smart TVs, routers and gaming consoles that use embedded memory.
Indian PC and phone brands sourcing components on tighter contracts, rather than spot-market deals, will feel this later but harder — many component contracts for 2026 were signed before the scale of the memory chip shortage became clear, so renewal cycles in 2027 could see sharper one-time price jumps for Indian buyers.
FAQ
Who is the Micron CEO who made this statement?
Sanjay Mehrotra is Micron Technology’s president and chief executive officer. He has led the company since 2017 and previously co-founded SanDisk.
Why is there a memory chip shortage in 2026?
Demand for AI data centre hardware, particularly high-bandwidth memory used alongside GPUs, has pulled wafer capacity away from everyday DRAM and NAND, while the industry simultaneously transitions from DDR4 to DDR5.
Will RAM and SSD prices fall after 2026?
Based on Micron’s own guidance, no — the company expects the memory chip shortage to be more severe, not less, in 2027 and 2028, since new fab capacity takes years to come online.
Does the memory chip shortage affect smartphone and laptop prices in India?
Yes. India imports the vast majority of its finished chips and devices, so global DRAM and NAND price increases typically reach Indian retail prices within one or two quarters.
Does Micron manufacture memory chips in India?
Not yet. Micron’s Sanand, Gujarat facility handles assembly, testing, marking and packaging of chips — the actual silicon wafers are still fabricated at Micron’s plants overseas.
Conclusion
Micron’s CEO isn’t hedging: the memory chip shortage that’s already pushing up prices in 2026 is set to deepen through 2027 and 2028, mainly because AI demand for high-bandwidth memory is outpacing new fab capacity that takes years to build. For India, that means costlier electronics are likely here to stay for a while, even as the country builds its own foothold in chip packaging through plants like Micron’s Sanand unit.