Mechanical hard drives are no longer in vogue as the primary storage medium for gaming PCs, but bulk HDDs still have their place. Primarily, they remain the best option for bulk backups. They can also help manage an ever-growing gaming library when SSD space is in short supply. This is why we always look forward to the detailed reliability reports that Backblaze, a cloud storage provider, publishes on the HDDs they use. In their latest report, Backblaze provides preliminary (and encouraging) data on the dozens of 16TB HDDs they use. [The larger the HDD, the more space there is for things you don't want to clutter up your SSDs, but the larger the HDD, the more space there is, because if a huge-capacity HDD fails, you could lose a lot of data at once.

Another option is to install multiple small-capacity HDDs, but that would undermine the convenience of a large-capacity storage medium in the first place. It is a matter of choosing your poison.

Another consideration is brand. Anecdotally, I have seen more people complain about Seagate HDDs than any other manufacturer. But are Seagate drives less reliable than other brands?

Backblaze operates storage pods ranging from 4TB to 16TB, including over 129,000 HDDs from various manufacturers. This cloud backup service regularly posts detailed statistics on failure rates, including an estimated "annualized failure rate." This is based on the formula (number of drive failures / number of drive days / 366) x 100. In non-leap years, 366 is replaced by 365.

In the most recent quarter, Backblaze added dozens of 16TB models to the mix; Backblaze only provides statistics on Seagate's 16TB drives because it excludes models used simply for testing purposes or with fewer than 60 drives in use Backblaze only provides statistics on Seagate's 16TB drives, as it excludes models used merely for testing purposes and models with fewer than 60 drives in use.

Here are the results:

The "Drive Days" category shows how many drives were used each day during a given period. For example, if 10,000 drives were used on a given day, the number of drive days for that day would be 10,000. Also, if these drives were used every day for 30 days, that would be 300,000 drive days for that particular model.

For the Exos X16, Seagate's 16TB enterprise HDD, Backblaze has data for 5,036 drive days based on this formula. More importantly, none of these drives have failed in the past three months.

This is encouraging, although limited in scope. Of the models recorded, the Exos 16 HDDS had fewer failures than any of the other drives; using Backblaze's calculation methodology, if one of these drives had failed in the past three months, the annual failure rate would equate to 7.25%.

In other words, there is room for interpretation in the data and Backblaze's calculation method: the HGST drives appear to be performing well, with an overall low failure rate; the HGST drives are fewer overall than the Seagate models, but from 4 TB to 12 TB of 29,074 drives, only 27 have failed in the past three months.

While none of the data is perfect, we have yet to find any further data on HDD reliability.