Don't Buy A Hard Drive Until You've Read This.


Hard Drive brand failure rateSo it's just after Christmas and your PC is so full of video and photos that you're thinking of upgrading your hard disk. Or maybe you still have a bit of cash left to spend in the sales, and you want to replace your existing mechanical drive (so-called "spinning rust") with an SSD.

But where do you start? You probably know what physical size of drive you need (2.5 or 3.5 inch, pretty much), and the capacity (anything from a few hundred megabytes to 12 terabytes). But which manufacturers' drives are the most reliable, and which models are worth paying slightly more for?

Ultimately, with hard disks it's all about the reliability. And if you want to find out about reliability, then you need to speak to someone who's used a particular drive for a while and ask them whether it's broken yet. Even better, you need to speak to an organization that has about 90,000 drives in operation, and look at all their stats about which drives are still working and how many have broken.

And, as luck would have it, you can do just that, through the excellent reports that are put out by a company called Backblaze. They're a US-based cloud backup provider, which is why they have 90,000 drives. And they produce regular reports, broken down by manufacturer and model number, on which drives are working and which aren't. And the latest figures are fascinating, as they reveal that one particular manufacturer's drives seem pretty unreliable, with a 31% failure rate.

Check out all the details for yourself, before you even think about buying a new disk (spinning rust or SSD). The best place to start is and all the information is free to access. You can even download some very detailed data if you want to do your own analysis.

Please rate this article: 

Your rating: None
Average: 4.5 (52 votes)


The Backblaze data comes from the biggest drives available, so it's not relevant to those of us to whom a 1T or 2T drive will be a step up.

Never ever buy a "factory refurbished" hard drive no matter what the brand. Worst money I've ever saved.

The Backblaze site is a useful resource but the data in this table is of very limited use without a lot more information. The annualized failure rate does not provide a fair comparison because it doesn't tell us anything about what really happened. I'd be looking at other issues.

Also, it's not correct to say that "one particular manufacturer's drives seem pretty unreliable, with a 31% failure rate." It was only one particular model of drive and not the entire range.

Lifetime failure rates. Drives tend to fail in large proportions at the same time in the life of the drive - usually early on or late in the lifetime. Many of the drives with low annualized failure rates will have higher failure rates at the end of their useful life. These Seagate drives are old so we would expect higher failure rates. As Backblaze themselves say "in Q1 the 4 TB Seagate drive model: ST4000DX000, has a high failure rate of 35.88%, while the lifetime annualized failure rate for this model is much lower, 7.50%."

They also say, "In this case, we only have a 170 drives of this particular drive model, so the failure rate is not statistically significant, ..."

Technology. The expensive Seagate drives are often pushing the technology with features that other drives don't have. The model in question is apparently unusual with an extra platter.

Usage. The drives might have been used quite differently than most of the other drives. Drives that are mainly read last a lot longer than drives that have heavy write loads. That is a more likely scenario because the Seagate model was an expensive Pro drive. Backblaze might have been using the advanced technology for activities that stressed these drives more.

I can't imagine that they would be worth buying for any retail consumer. They didn't even get to any retail outlet in my country.

Sourcing: Are all the drives failing from the same manufacturing batch? Given the small number (less than 200) Backblaze were using then they might have bought them as one small lot. So Backblaze might have picked up most of the bad drives in that model.

Cost: It is price-performance that is usually important. It is not as simple as saying "Ultimately, with hard disks it's all about the reliability." and more so when the reliability figure you're quoting is so likely to be misleading.

(edited to correct mispelling and bad grammar)

Good post Remah :)

Thank you! Your comment is very informative.

Thanks for this article. I know Seagate always gets a bad rap. This is anecdotal, but I have always had a good experience with Seagate drives. We have 5 in use right now and have owned 5 or 6 others over the past 10-15 years. My bad experiences have been with the two WDC drives I bought. Go figure!

I notice even though Seagate gets the lowest marks, Backblaze just installed another 1200 of them.

Same here. I have had good experience with Seagate drives. I have a Seagate drive in desktop, and an external hard drive... both working fine so far, even after several years.

Same for me. I have a Seagate Expansion Drive that gets all kinds of stick including moving daily between other PC's and the TV. Three years now and never a glitch. MC - Site Manager.

Glad to hear of other happy Seagate users. Here are my current stats for all my local drives:
D: Seagate ST4000VN000 (internal) power on time (pot) 859 days, 100% health (per Hard Disk Sentinal), heavy use;
Q: Seagate ST4000LM016 (external) pot 715 days, 100% health, light use;
K: Seagate ST8000AS0002 (external), pot 467 days, 100% health (but runs hot), daily backups;
G: Seagate ST5000DM000 (external), pot 471 days, 100% health, light use;
W: WDC WD4001FAEX (internal), pot 1269 days, 100% health (but runs noisy and has always made clicking noises which made me nervous), now light use, was heavy the first year or two
C: Samsung SSD 840 EVO (internal), pot reports as 1768 days which is incorrect - I haven't owned it that long, 91% health, heavy use (boot drive)

Very interesting. Good Article.