SSD vs HDD price per TB: is flash catching up?

Every few months someone declares that SSDs are about to make NAS hard drives obsolete. The logic sounds right: flash keeps getting cheaper, big hard drives take forever to resilver, and nobody's saturating a spinning disk anyway. So I pulled my own price history to check whether the one number that actually decides it, price per TB, is really closing.

It isn't. Not yet, and not close.

Both lines are the monthly median price per TB across the drives I track (hard drives 8TB and up, SSDs 1TB and up, US pricing, from daily price history). Two things stand out.

First, the gap has been remarkably steady. For about a year and a half the typical hard drive sat near $20 per TB and the typical SSD near $120, a gap of roughly five times. Even if you ignore the median and line up the cheapest hard drive against the cheapest value SSD, it's still three to four times. That ratio has barely moved in two years. For flash to actually displace disk on price, that line has to trend down, and it isn't.

Second, the 2026 storage crunch hit both, but it hit flash harder. When NAND and memory prices spiked, SSD price per TB roughly tripled while hard drives roughly doubled, so for a while the gap got wider, not narrower. Flash has since recovered faster than disk, which pulled the ratio back to about where it started. Over the whole period it's a flat gap with a spike in the middle, not a convergence.

So what does that mean if you're building or expanding a NAS?

For bulk storage, hard drives still win by a mile. Media libraries, backups, archives, anything where you mostly care about capacity per dollar, a 5x price gap is not a close call. A pool of large hard drives is still the obvious choice, and the manufacturers clearly agree: the roadmap runs to 40TB and beyond on HAMR, which isn't what a dying product looks like.

Where SSDs already win is where they always have: speed, and the things that need it. Metadata and special vdevs, a read cache, small-file or VM datasets, an all-flash pool for one fast workload. That's the real shift happening now, and it's a hybrid one: SSD for the performance tier, hard drives for the bulk tier, in the same box.

The real pressure on big hard drives is not price, it's rebuild time. A 30TB drive takes a long time to resilver, and you spend that whole window degraded. That pushes you toward RAIDZ2 or Z3, narrower vdevs, or dRAID, not toward replacing the disks with flash you can't afford at that capacity.

The honest summary: SSDs are not about to take over NAS bulk storage, because the price gap that would have to close has stayed open at around five times for two years. The flash transition is real, but it's slow and hybrid, measured in many years rather than one or two. When the gap does start to close for real, this chart will show it first, because it updates from live pricing every month.

The whole dataset behind this is open (CC BY) with a JSON/CSV API if you want to run the numbers yourself.