Which is better - Seagate, WD or HGST

This has been a argument for a long time, I have drives from all three and have had no issues apart from 1 of three 1TB drives bought at same time arriving dead. That just happens so I keep on buying the drive from respectable manufacturer on price per GB. Lately it has been Seagate with that win.

http://www.extremetech.com/computing/198154-2014-hard-drive-failure-rates-point-to-clear-winners-and-losers-but-is-the-data-good

Judge for yourself

the guy above recommending a Seagate clearly doesn't know that they are the ones to avoid.
Personally i would go for a WD black( top quality, very reliable)

Gamers like WD velo raptor which got 1000 rpm but if you can't get that get a WD caiver black it's a step down from WD raptor if you want speed go with a ssd if you want space + speed go with hybrid

instead of looking a a few graphs you might want to consider reading the whole article.paints a very different picture.you might also want to check into rollis record here at toms.personally,and this is all your gonna get is ive had not trouble with seagate,wd,maxtor or any other hdd.when asking the question,which is better,your going to get mostly personal opinion.

The following represents % of warranty returns between 6 and 12 months; each listing has the just completed study period followed by the preceding 6 months.:

Ending 04-30-2014
http://www.hardware.fr/articles/920-6/disques-durs.html

- Seagate 0,86% (contre 0,95%)
- Toshiba 1,02% (contre 1,54%)
- Hitachi 1,08% (contre 1,16%)
- Western 1,13% (contre 1,19%)

Click to expand...

Ending 11-06-2014
http://www.hardware.fr/articles/927-6/disques-durs.html

- Seagate 0,69% (contre 0,86%)
- Western 0,93 (contre 1,13%)
- HGST 1,01% (contre 1,08%)
- Toshiba 1,29% (contre 1,02%)

Click to expand...

Ending 05-19-2015
http://www.hardware.fr/articles/934-6/disques-durs.html

- Seagate 0,68% (contre 0,69%)
- Western 1,09% (contre 0,93%)
- HGST 1,16% (contre 1,01%)
- Toshiba 1,34% (contre 1,29%)

Click to expand...

What is obvious from the above is that HD failure rates are very low if you exclude those that are DOA. It's important to note that failure rates increase w/ larger size.... I pretty much don't take individual drive failure rates into consideration as long as they don't exceed 2%.

One final note on the numbers is that usually you see the rankings shuffle a bit as the result of the proverbial "bad eggs" that make it to market now and then, It is significant therefore to note that the relative rankings remained virtually unchanged for 4 consecutive periods.

When looking at such studies it is important to note the context.... one example is the infamous and ridiculous Backblaze study which peeps like to post to support an argument w/o bothering to read the test parameters. They used a bunch of consumer drives in a server environment; consumer drives are equipped with various protection features (i.e. head parking) which protect the drives from damage by parking the arm / heads when drive is not in use. Drives are rated for 250,000 to 500,000 cycles which is well beyond what consumer drives see in typical usage. However in a server environment, they can burn 50,000 cycles in a month which leads to premature failure.

The drives that do well under the Backblaze conditions, serve poorly in a consumer environment. Because of the literally rock-solid building design for server farms like used for the Backblaze study, drives designed for server use need not be equipped with anti-vibration and head parking features because they are unnecessary. Put them in a consumer desktop and one "desk bump" by a family member / pet can crash the heads killing the drive.

What can we take away from the Backblaze study .... don't use consumer drives in a server environment and don't use server drives in a Home PC / Office Workstation. More importantly, don't rely on data from PC components being used in an inappropriate manner.

From anadtech site

Although I do appriciate the backblaze study, they make it CLEAR their failure rates for all the drives ranked are for a SPECIFIC application, one that none of the drives are designed for (these are consumer drives running in a cold-storage scenerio among other drives in RAID.) Almost none of those drives have firmware that supports staggared spinup, vibration monitoring, harmonic balancing, differential queuing, and so on.

To the point: Seagate consumer drives since the 7200.10-12 up to the Cuda' XT have very aggressive head parking (called load cycles) and they are rated at 250,000-500,000 load/unloads. I've seen drives rack up 90,000+ load/unloads in months if you use them for heavy access (seeding torrents.) NAS\RAID drives have firmware that often completely disables head parking (smart powersaving parameter 0xC1:255)

The reason the Hitachi drives do so well in the Backblaze study is because Hitachi never even implemented the SMART 0xC1 command.

Click to expand...

FYI, it's Seagate, WD, and Toshiba now. HGST merged with WD, but as a condition of their merger regulators required HGST to sell its 3.5" manufacturing assets to Toshiba (which used to only make 2.5" drives). So the three remaining HDD manufacturers are Seagate, WD (with HGST's 2.5" manufacturing), and Toshiba (with HGST's 3.5" manufacturing).

There is no "best" manufacturer. Each different model drive varies in its reliability, and all manufacturers have put out good models and bad models. The manufacturers try to implement new, denser storage technology and cheaper manufacturing with each new model. Sometimes it works, sometimes it doesn't. Statistically, you can't really tell if a particular model is good or bad until 2-3 years after it's been out and enough drives have failed to calculate reliability with a margin of error smaller than the failure rate. So realistically it's impossible to know ahead of time if you're getting a good or bad model (unless you like buying old drives which are 3+ years old). Just buy whichever brand makes you comfortable or whose price you like, and be diligent about making backups.

1. It is significant and appropriate that HGST be listed separately as even tho they are an owned subsidiary of WD, HGST branded dives are still manufactured in HGST fabs. AFAIK, the original terms of the agreement remain in place and they remain in fact "competitors". Another part of the agreement required WD to divest of assets which would thereby allow Toshiba to make entry into the 3.5" desktop market

https://en.wikipedia.org/wiki/HGST

" It was agreed that WD would operate with WD Technologies and HGST as wholly owned subsidiaries and they would compete in the marketplace with separate brands and product lines"

2. You can tell statistically as early as 6-12 months out. They don't wait to years to do MTBFs. When you see numbers like this .... just 6 - 12 months into a production run, these drives are best avoided.

- 4,76% WD Black WD4001FAEX
- 4,24% WD Black WD3001FAEX
- 3,83% WD SE WD3000F9YZ
- 2,56% HGST Travelstar 7K1000
- 2,39% Toshiba DT01ACA300

Again, that shouldn't be taken to mean that "WD is the worst".... all it can be taken to mean is that at the time 3 of their larger 3 - 4 TB models have some issues. Six months later it is entirely possible that improvements will be made and 5 different drives mate the Top 5 Razzie Awards

- 4,58% WD Red WD60EFRX
- 3,40% Toshiba 3 To DT01ACA300
- 2,93% WD Green 4 To WD40EZRX
- 2,78% WD SE 3 To WD3000F9YZ
- 2,14% Hitachi Ultrastar A7K2000 1 To

Notice that the 2 Blacks have improved ... the WD4001FAEX (4,76%) has been replaced by the WD Black WD4003FZEX (1,18% ) w/ less than 1/3 the failure rate. And the WD SE WD3000F9YZ that was at 3,83% has dropped to 2,78%. But, going the other way the, Toshiba DT01ACA300 went from 2.39% to 3.40%.

I haven't checked Toshiba's latest drives, but their 3.5" drives from late last year were quite obviously rebranded HGST drives.

2. You can tell statistically as early as 6-12 months out. They don't wait to years to do MTBFs. When you see numbers like this .... just 6 - 12 months into a production run, these drives are best avoided.

- 4,76% WD Black WD4001FAEX
- 4,24% WD Black WD3001FAEX
- 3,83% WD SE WD3000F9YZ
- 2,56% HGST Travelstar 7K1000
- 2,39% Toshiba DT01ACA300

Click to expand...

Anyone can write down numbers to 0.01%. That doesn't mean it's accurate to 0.01%.

For a 95% statistical certainty with a margin of error of 0.01% at a 3% rate, you'd need a sample size of 11 million. The manufacturer is unlikely to get a sample size that large in 6-12 months, much less a single store. Even a margin of error of 0.1% requires a sample of 111,800. Per drive model. Which again is unlikely for a single store. So basically, the two numbers after the decimal point in your figures are meaningless.

A margin of error of 0.5% (3% +/- .5% - meaning you can just barely detect a difference between 2%, 3%, and 4%) requires a sample of 4472 per model. A retail store might be able to get that in a year. More than likely it's down near 1000 units/yr for each model, which would give it a margin of error larger than 1% (i.e. 2% is not distinguishable from 4%). So like I said, it will take several years of data to determine with some confidence if a particular drive model is more or less reliable than others.

Margin of error with a 95% confidence interval for a measured 3% failure rate is 1.96*sqrt [ (3%)*(100%-3%) / n ]. This is the loosest standard for drawing conclusions used in science and statistics.
For a 99% confidence interval, margin of error is 2.58*sqrt [ (3%)*(100%-3%) / n ]