New HDD - Toshiba 16 TB - But Backblaze article came out = Bad drive...?


Mar 14, 2017

Im in the process of choosing a new HDD, i had some drives in mind, but on some forum i was recommended even bigger and "better" drive for the same price - Toshiba 16TB, MG08ACA16TE. I liked it a lot from all the descriptions... But just few days ago new Back Blaze article about their HDD failure rates for 2022 came out

And I by pure accident came across it. What was my surprise when i saw that my new choice is among them - MG08ACA16TE.

And it doesnt look good... Although the drive SHOULD be GOOD (has 5 years warranty, 2.5 milion MTBF etc...) , BUT... according to the REAL WORLD samples/stats, it doesnt really look good... They have like 6000 of these (!) so the numbers that they give should give you a good idea about that HDD in general (if they had only 2 drives, who cares, 6000 drives is according to me a good sample size to make a judgement). They ran only for like a year (14,7 months) and ALREADY (!) they have AFR of 1,57% (!). Thats like almost the worst (if not WORST) number of all the drives with similiar "age" they have...

According to this:

The AFR should be (is stated that it is): 0,35%... And JUST in the 1st year of REAL WORLD USAGE (with 6000 samples (!) ) showed the real world AFR of 1,57%, thats like 4.5 times more than it should be...
So... this drives appears to be a BAD drive... right? And i should NOT go for it in light on these newly revealed real life "tests"...? Or is it possible that this drive just fails more in the first year and THAN it stabilizes its AFR on a much better number...? According to my understanding, if a drive exhibits unexpectedly high failure rate for the 1st year, its safe to assume, the failure rate will NOT get any better with years passing...?


What would you thought about these revelations about the drive...? From all i know about PC HW and HDDs it appears to me, that this is a bad drive... (?) (in terms of reliability)

Thank you
Your 1.57% AFR is absolutely nothing to worry about.

Hard drives, like many devices with moving parts or parts with cyclical heating/cooling cycles, operate on the 'bathtub curve'. The bathtub curve is that they tend to fail very early in life, or very late in life, and be very reliable in the middle. These drives at Backblaze will be relatively early in their lifespan. If you want to look at the Backblaze stats and find a drive model to worry about, it would be the ST14000NM0138 with it's 5.7% AFR despite being barely older than the Toshiba drive.

Also, you describe the 1.57% AFR as if it is unique or high on the results. Overall the AFR for all drives in 2022 is 1.37%, so your 1.57% is barely above that. And on the chart there are nearly 67k drives with a higher AFR than your Toshiba model.

Regardless of all of that, you're talking about buying a single drive or a small quantity of drives. Statistically speaking, if *any* drive fails on you it would qualify as a statistical anomaly. Even with the Seagate's 5% AFR, that's still a 95% chance (overwhelming) that an individual drive does *not* die on you in a given year. I'm also willing to bet that Backblaze works their drives significantly harder than you will. Basically with a sample size of 1 or 2 drives, it comes down to luck not statistics.

Here's the point of all of this: everything can fail, but nothing is likely to fail. If you care about the data you are putting on this drive, then you need to back it up. Preferably that backup would be kept offline, or in a different location (ala Backblaze) or a NAS or something. You either care about the data - so you should back it up - or you don't care about the data and failures are just an unlikely inconvenience. You'll have to make your choices from there!