The industry is preparing to make more drives, not just bigger ones.
In the quiet hum of clean rooms and semiconductor-like fabrication lines, the hard disk drive industry is making a wager on the future of artificial intelligence infrastructure. Seagate, TDK, and Western Digital are expanding their manufacturing capacity for magnetic heads — the critical bottleneck in drive production — signaling a belief that the world's appetite for dense, affordable storage will only deepen through the end of the decade. After years of retreat into a narrower, higher-capacity niche, these manufacturers are now betting they can grow both the size and the number of drives they ship. The factories being built today will answer, by 2027 or 2028, whether that conviction was wisdom or overreach.
- After 18 months of obscuring unit shipment data behind raw capacity figures, HDD manufacturers are now signaling through factory expansions that volume growth — not just bigger drives — is back on the table.
- Seagate is nearly tripling its Bloomington, Minnesota clean room from 11,000 to 19,000 square feet, a concrete and costly commitment to HAMR head production that cannot easily be walked back.
- TDK's parallel April announcement of increased head production capacity confirms this is an industry-wide posture shift, not a single company's gamble.
- Analyst projections now span an 11–30% range in unit shipment growth from 2026 to 2030 — a gap wide enough to represent two entirely different futures for the industry.
- The new facilities won't come online until late 2027 or 2028, meaning the industry must sustain its conviction through years of uncertainty before the bet is settled.
The hard disk drive industry has spent the past year and a half presenting itself in carefully chosen terms — releasing storage capacity figures while staying quiet about how many actual drives were shipping. That selective transparency is now giving way to something more legible: factory announcements that suggest manufacturers expect to be building a great many more drives, not merely larger ones.
Seagate's expansion is the clearest statement of intent. The company is nearly tripling its clean room space in Bloomington, Minnesota, from 11,000 to 19,000 square feet, with the new capacity oriented toward HAMR heads — components that use integrated lasers to enable denser recording. Magnetic heads are the production bottleneck; you need one per platter, and they're fabricated through processes that resemble semiconductor manufacturing. More head capacity means more drives. TDK, which supplies heads to Toshiba and also serves Seagate and Western Digital, made a similar announcement in April. Western Digital, for its part, has been engineering its way toward higher-component drives — moving from 10-disk to 14-disk configurations — a path that implies its own head capacity pressures, even if no formal expansion has been announced.
The numbers behind these moves are striking. HDD unit shipments grew just 0.4% from 2024 to 2025, and only 1.2% growth is projected for the following year. But Coughlin Associates, which tracks the industry closely, now projects unit shipment growth of 11% to 30% between 2026 and 2030. The lower figure represents a baseline; the higher one reflects the optimistic scenario these factory expansions appear to endorse.
What changed the calculus is artificial intelligence. Data centers running AI workloads require enormous volumes of storage, and hard drives remain the most economical solution at scale. For years, as solid-state storage eroded the consumer and enterprise HDD market, manufacturers narrowed their focus to fewer, bigger drives. Now they appear to believe that AI demand is durable enough to justify growing both dimensions at once — more storage per drive and more drives overall.
The new facilities are expected to come online in late 2027 or 2028. Between now and then, the industry will have committed capital and clean room square footage to a vision of sustained demand. Whether that vision proves accurate is the question these expansions have placed on the table.
The hard disk drive industry is quietly preparing for a significant shift. For the past year and a half, the major manufacturers have been cagey about how many drives they're actually shipping—they've released storage capacity numbers instead, which obscures the real picture. But a series of recent announcements about factory expansions suggests that's about to change, and that the industry expects to be making a lot more drives, not just bigger ones.
Seagate's move is the clearest signal. The company is nearly tripling the size of its hard disk head manufacturing facility in Bloomington, Minnesota, expanding from 11,000 square feet of clean room space to 19,000 square feet. Magnetic heads are the critical bottleneck in drive production—they're manufactured on ceramic wafers using semiconductor-like processes, and you need one for every platter in a drive. The new capacity will likely focus on HAMR heads, which use integrated lasers to heat the recording media and allow for denser data storage. When Seagate discussed these plans with investors, the company said it was prioritizing capacity shipments over unit volume. This expansion suggests a change in that calculus.
The timing matters. Seagate's new manufacturing space should be operational in roughly 12 to 18 months, which means additional production capacity arriving in late 2027 or 2028. That's not immediate, but it's concrete. Western Digital, meanwhile, has been taking a different approach—packing more disks and heads into each drive, moving from 10-disk configurations toward 14-disk models. The company hasn't announced head capacity expansions, but the logic suggests it should, both to support those higher-component drives and to manufacture more of them.
TDK, which manufactures heads for Toshiba and supplies Seagate and Western Digital as well, announced in April that it too was increasing head production capacity. The pattern is unmistakable: the industry is preparing to make more drives.
What's driving this? The numbers tell the story. From 2024 to 2025, hard disk drive unit shipments grew by only 0.4 percent. From 2025 to 2026, the projection is a modest 1.2 percent increase. But analysts at Coughlin Associates, which tracks the industry closely, now project that unit shipments could grow anywhere from 11 percent to 30 percent between 2026 and 2030, depending on how aggressively manufacturers pursue volume. The baseline scenario assumes 11 percent growth. The optimistic scenario—the one these capacity announcements seem to support—assumes 30 percent.
The difference between those two futures is substantial. It hinges on whether the industry decides to chase unit volume or continue focusing narrowly on storage capacity. For years, as solid-state storage ate into the HDD market, manufacturers retreated into a narrower business: making fewer, bigger drives for data centers. But artificial intelligence workloads are changing the equation. Data centers need massive amounts of storage, and hard drives remain the most cost-effective way to provide it at scale. The capacity expansions suggest manufacturers believe that demand will be sustained, and that they can grow both their storage shipments and their unit shipments simultaneously.
The real test comes in 2027 and 2028, when these new facilities come online. If demand materializes as expected, the hard disk drive industry will have positioned itself to capture it. If it doesn't, the manufacturers will have built capacity they don't need. For now, the bets are being placed.
Citações Notáveis
Seagate previously told investors it was prioritizing storage capacity shipments over unit volume, but the facility expansion suggests a shift in strategy.— Industry analysis based on investor communications
A Conversa do Hearth Outra perspectiva sobre a história
Why does it matter whether they're shipping more drives versus bigger drives?
Because it tells you what they believe about the future. If you're just making bigger drives, you're optimizing for a shrinking market—fewer customers, each needing more storage. But if you're expanding the factories that make the components, you're betting that you'll have more customers, period.
And these head manufacturing facilities are really the constraint?
They're one of the main ones. You can't make a drive without heads. They're precision components, manufactured in clean rooms like semiconductors. If you want to double your drive production, you need to roughly double your head production capacity. It's not something you can quickly improvise.
So Seagate's expansion means they're confident about demand 18 months from now?
It means they're confident enough to commit capital to it. But there's also a lag built in—they're not saying they'll ship more drives next quarter. They're saying that when this facility is ready, in late 2027 or 2028, they'll be able to. That's a long bet.
What changed? Why are they suddenly interested in unit volume again?
AI data centers. They need enormous amounts of storage, and hard drives are still the cheapest way to provide it. SSDs are faster but much more expensive per terabyte. For cold storage, archival, the kind of thing that powers large language models, hard drives make economic sense at scale.
But Western Digital isn't expanding head capacity, you said?
Not that we know of. But they're doing something different—packing more heads and disks into each drive. That's another way to increase capacity without necessarily increasing unit volume. Though logically, if they want to make more of those high-capacity drives, they'll need more heads eventually.
So the industry could grow 11 percent or 30 percent in the next four years?
That's the range. The baseline assumes modest growth. The optimistic case assumes these expansions translate into real demand and real production. The announcements suggest manufacturers are betting on something closer to the optimistic case.