What better way to find out which hard drives hold up the best than to ask a company that uses lots of them?
Cloud backup firm Backblaze uses more than 27,000 consumer-grade hard drives in its data centers, which prompted a question from one of its users about which drive makers produced the best hardware. After crunching the lifetime and failure stats of the drives in its inventory, Backblaze published the results in a blog post -- and the differences between the best and worst consumer-grade drives, in terms of longevity and durability, were striking.
The best drives Backblaze currently use are the Hitachi Deskstar models, which sport failure rates as low as 0.8 percent and as high as 2.9 percent annually. If you're curious, the exact drive models at the low end of the failure curve are the Hitachi Deskstar 5K3000 and 7K3000, both 3TB drives.
The worst -- by a wide margin -- were the Seagate Barracuda 7200 models, 1.5TB drives that died at a stunning rate of 25.4 percent annually.
This isn't to say all of Seagate's drives were duds. Its Seagate Desktop HDD.15, a 4TB drive, has only racked up a 3.8 percent annual failure rate so far. But on the whole, Seagate's drives failed that much more quickly than the competition -- and the smaller-capacity drives, the 1.5TB and 3TB models, failed more often.
Western Digital's drives generally did much better, with the high end of its annual failure rates around 3.6 percent. But the overall winners were drives by Hitachi. The Hitachi drives with the worst failure rates, the company's 4TB models, only died around half as often as the best Western Digital drives.
Each brand of drive also had different failure profiles, according to the stats Backblaze collected over the course of three years. Hitachi drives were the most consistently reliable. Western Digital drives tended to have most of their failures up front, within the first couple of months of use; those that survived lasted. Seagate drives, on the other hand, "die off at a consistently higher rate, with a burst of deaths near the 20-month mark," wrote Backblaze's Brian Beach.
An important note to keep in mind with these statistics is that they can't be used to predict the lifetime of any individual drive. They're useful for gauging the quality of a given make or model of drive overall, but even with these numbers, it's still entirely possible to buy a dud Hitachi drive and follow it up with a long-lived Seagate. Richard Elling of Oracle once blogged about how the widely misunderstood issue of MTBF (mean time between failure) can affect systems design -- including, naturally, a system as large as Backblaze's.
This story, "Buying a hard drive? Stats say get Hitachi, avoid Seagate," was originally published at InfoWorld.com. Get the first word on what the important tech news really means with the InfoWorld Tech Watch blog. For the latest developments in business technology news, follow InfoWorld.com on Twitter.