External Drives Have Caught Up With Bare Drives Per Terabyte

The premium that defined this category has gone

For most of the last decade, buying storage came with a rule so consistent that people stopped checking it: the drive inside an external enclosure cost less to buy on its own. The enclosure, the USB controller, the power supply and the packaging were a premium you paid for convenience, and the whole practice of shucking existed to capture that premium back on the occasions a sale went deep enough to invert it.

Our own tracking says that rule has stopped describing the market. The comparison block above is computed when this page loads, from in-stock listings, and it puts the cheapest external desktop drive per terabyte alongside the cheapest bare internal SATA drive per terabyte at the same condition.

On new listings the two are effectively level — within a couple of percent of each other, which is inside the noise of a single promotion on either side. On used listings the external is ahead outright. Neither of those is the dramatic inversion that a headline would prefer, and the convergence is the more interesting finding, because it is structural rather than promotional.

We are stating this carefully on purpose, because there is a version of this story circulating with much bigger numbers attached, and we could not reproduce it.

What we could not reproduce, and why we are saying so

The claim that prompted us to run this analysis was that external desktop drives had reached rates below ten dollars per terabyte at the top capacities, comfortably undercutting bare drives. We went looking for that in our own catalogue and did not find it.

The largest external in our tracked stock is 24TB. There is no 26TB external in our tables at all, at any price, and the cheapest external of any capacity or condition sits well above ten dollars per terabyte. If such a listing existed in the venue we track, at the moment we looked, we would have it.

That is not a claim that the reported figure never happened. Externals run deeper and more frequent promotions than any other storage category, a single retailer's doorbuster at a single capacity can produce a startling rate for a few hours, and we price one venue rather than the whole market. A number that was true somewhere on a Tuesday is not a number we can show you today.

What we can show you is what is buyable right now, which is the only figure that helps somebody deciding what to buy. Our tables exclude out-of-stock listings entirely rather than carrying them at a last-known price, and that discipline is why our numbers are less exciting and more actionable than the ones being repeated.

There is cheaper storage, and it comes with a catch

One more correction is owed before the analysis, because it is the sort of thing a page like this usually skips.

External desktop drives are not the cheapest storage per terabyte in our catalogue outright. That title belongs to second-hand enterprise SAS drives at small capacities, which sit meaningfully below anything in the external tables. Those drives are real, they are buyable, and we track them.

The catch is what they need around them. A SAS drive does not plug into an ordinary computer. It needs a host bus adapter, a machine with a slot to put it in, and cabling that consumer motherboards do not provide. For somebody with a server and a spare PCIe slot that is a small obstacle and the rate is genuinely the best available, which is why our cheapest per TB table ranks them where they fall.

For everybody else it is a different product category wearing a similar price tag. So the accurate claim, and the one we make on our external drives hub, is narrower and more useful: external desktop drives are the cheapest storage per terabyte you can plug straight into an ordinary machine and use.

Why the enclosure stopped costing anything

The convergence is not the result of enclosures becoming cheap to make. Cases and USB controllers were never the expensive part.

What changed is upstream. The drives that go into retail external enclosures and the drives that reach retail shelves as bare units move through different supply arrangements, and the 2026 shortage has priced those arrangements differently. Hyperscale demand absorbed nearline production and pushed bare drive pricing up hard, while the consumer external channel — planned further ahead, contracted differently, and aimed at a price-sensitive buyer who walks away — moved less.

The result is that the retail bare drive lost its structural advantage rather than the external gaining one. That distinction matters for anyone predicting what happens next: this is a shortage artefact, not a permanent repricing of what an enclosure is worth, and the historical ordering could return when nearline supply loosens.

It also explains why the used comparison looks different from the new one. A used bare drive in our catalogue is usually a decommissioned enterprise pull with a published part number and a known workload history. A used external is a consumer unit whose insides were never published even when it was new. Those two things trading at similar rates is not one product beating another; it is two different markets that happen to intersect on one metric.

What this does to the case for shucking

Shucking — buying an external and removing the drive — exists entirely because of the gap this article is about. If the gap closes, the practice stops making sense, and that is roughly what has happened at some capacities and emphatically not at others.

The honest answer is that shucking is now a per-capacity question rather than a general strategy. Our external drives hub computes the cheapest external at each capacity against the cheapest bare internal SATA drive at the same capacity, on new listings, and the table has rows going both ways. At some capacities the external is meaningfully cheaper and the trade is worth making. At others the bare drive already costs less, and opening an external there would cost you money and void a warranty in exchange for nothing at all.

That is a real change in how to approach this. The old advice — watch for a deep external sale, buy several, open them — assumed a persistent structural gap that a sale widened. The current advice is to check the specific capacity you want on the day you buy, because the sign of the difference is no longer predictable.

And the three costs of shucking have not moved at all. You still do not know which drive is inside before opening the case, and the contents still vary between production batches of the same model number. Opening still ends the warranty. And some units still use a 3.3V power-disable pin that stops the drive spinning up in a normal desktop power connector. Our shucking guide covers each of them properly.

A narrower gap makes all three matter more, because there is less saving to justify them with.

What nobody involved will tell you about the drive itself

There is a structural asymmetry between these two purchases that pricing analysis tends to skip, and it is arguably more consequential than the few percent this article has been measuring.

When you buy a bare drive you know what you are getting. The model number is published, the recording technology is published, the workload rating and warranty term are published, and you can look all of it up before you spend anything. Our CMR versus SMR list exists because those disclosures exist.

When you buy an external you know none of it. No manufacturer publishes which bare drive is inside an enclosure, at any capacity, from any brand. The contents vary between production batches of the same model and part number. A meaningful share contain shingled recording, which is a perfectly reasonable choice for the job an external actually does and close to disqualifying for an array.

That is why every external row on our site reads "not disclosed" in the recording column and will keep reading it. Guides that print the contents as fact are reporting community teardowns — accurate accounts of particular units — and quietly converting a batch observation into a product specification.

So the comparison this article makes is genuinely about price and only about price. At parity per terabyte, the bare drive is the better-documented purchase by a wide margin, and that is worth something the tables do not capture.

What to actually do about it

If the storage is going to sit on a desk and plug into a USB port, buy the external. It is now competitive on the one metric that used to argue against it, it arrives ready to use, and the documentation you are giving up is documentation you were never going to need for that role.

If the storage is going into a machine's bays or an array, buy the bare drive. At current parity you are paying nothing extra for a published model number, a stated workload rating, a claimable warranty and a recording technology you can check before purchase. That was a trade worth considering when the external was clearly cheaper; at parity it is not a trade at all.

If you were planning to shuck, check the specific capacity before committing. The gap goes both ways now, and the version of this decision that used to be a reliable saving has become a calculation.

And if you are buying externals at all, the fraud risk in this category deserves more of your attention than the few percent this article has been about. Fake-capacity listings — a small flash chip with a controller reflashed to report a size it does not have — concentrate at large advertised capacities from unfamiliar brands, and they fail only after you write past the real capacity, which is usually after the return window. We reject any external HDD under eight dollars per terabyte and any unfamiliar brand above 2TB, at seeding and again when a page renders, and our external drives hub sets out what to look for if you shop elsewhere.