Towards the death of hard drives? No, we are moving towards the end of storage solely on flash, according to VDURA

One could quote Mark Twain when discussing the market for traditional hard drives: "the reports of my death are greatly exaggerated."

In fact, according to VDURA's CEO, Ken Claffey, we are actually seeing the opposite: the end of storage servers that rely solely on SSDs. What seemed like an inevitable evolution must now face the crisis surrounding solid-state memory, which includes NAND memory used in SSDs.

Are we at the end of "all-flash" servers? We quote Mark Twain, but Claffey has chosen instead to reference the famous tale of the emperor's new clothes in an article published on the company blog. Claffey highlights an open letter issued by Charles Giancarlo, CEO of the recently renamed Everpure, where Giancarlo explains why the company is raising prices for its products; Claffey notes that "[Giancarlo] refers to this as the third disruption of 'once every ten years' in the supply chain over the past five years."

Claffey continues by stating that the claims made by Everpure, VAST, and other entities regarding the affordability of building data storage systems based solely on flash memory have never been true. He adds: "this is not a criticism of flash memory. It is the right medium for 'hot' data to achieve performance, for metadata, for checkpoints, and it always will be. This is a critique of an architectural proposal that bets your entire data estate against a good that could never remain economical and which the people making the proposal never checked."

And it is precisely this aspect that has led to SSD prices now being over 20 times higher than traditional hard drives per terabyte of capacity and, consequently, systems using them to skyrocket in price. The issue, Claffey claims, is that oftentimes manufacturers of all-flash systems have made unrealistic claims, such as promising reductions in used space of 4, 5, or even 6 times due to deduplication and compression algorithms, and they used these to show how the cost of SSDs was not, in their view, that much higher than that of hard drives. Since the basic assumption was unrealistic, so too was the conclusion, and SSDs have always been more expensive than their spinning platter counterparts.

An interesting aspect pointed out by Claffey is that "the people building the largest data infrastructures in the world have never been fooled. Google’s Colossus is not all-flash. Meta’s storage backbone is not all-flash. Microsoft Azure’s storage is not all-flash. Amazon S3 is not all-flash. Each of them uses a mixed media architecture defined by software. Enough NVMe flash to saturate workloads and then HDDs for everything else that does not need the speed of flash memory."

The reason is that "they did the math. They realized that flash memory is a medium for performance, not for large volumes, and pricing your infrastructure based on a single good that you do not produce is a strategic mistake you can only make once."

The solution, according to Claffey, is to use systems like those from VDURA that natively manage different types of storage media in a unified manner. This way, one does not rely on a single component.

However, hard disks are also increasing in price.

There is, however, a note to be made regarding this discussion. While it is very true that the price of flash memory has skyrocketed between the end of 2025 and the beginning of 2026, it is also true that the price of traditional hard drives has significantly increased as well. Just the fact mentioned by Claffey that large hyperscalers still use HDDs has led to prices for these to surge, as production has already been booked: for example, WD stated that its entire production for 2026 has already been sold.

So while the lower prices of hard drives do provide companies with shelter from the shocks of the SSD market, on the other hand, the hard drives themselves are subject to significant price hikes. Architecturally and economically, Claffey is right (why keep all your data, even those not frequently used, on extremely expensive media?), but in the current situation, the choice is between absurd increases and disproportionate increases, if you will forgive our informality. It is difficult to determine what is better.