Search Amazon.com:
Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us $3,400 per month for bandwidth bills alone, and since we don't believe in shoving popup ads to our registered users, we try to make the money back through forum registrations.
«2 »
  • Post
  • Reply
Alereon
Feb 6, 2004

For me but LEFTHANDED

October 2013 hardware return rate statistics are out!

Every six months the French site Hardware.Fr publishes return rate statistics from an anonymous French merchant LDLC.com, the largest tech vendor in france (thanks Sh4!), listing the top brands in each product category by reliability, as well as the worst products in each category with the highest returns. Here's the Google Translate link to the latest issue, with details and analysis by section.

Motherboards
  1. Gigabyte 1.43%
  2. MSI 1.83%
  3. Asus 1.86%
  4. ASRock 2.09%
There is no longer any meaningful difference in reliability between motherboard brands. This is surprising to me because I expected to still see evidence of lower quality power delivery components on Gigabyte motherboards, but that does not appear to be the case.

Power Supplies
  1. Fortron/FSP 0.49%
  2. BeQuiet 0.61%
  3. Antec 1.33%
  4. Cooler Master 1.52%
  5. Seasonic 1.6%
  6. Thermaltake 1.87%
  7. Akasa 1.92%
  8. Corsair 1.96%
  9. Cougar 2.41%
There isn't a lot of spread among power supply manufacturers, and indeed, no power supply from a major brand had a return rate of 5% or higher during the period. The upshot seems to be that among the reputable brands there are no truly awful power supplies, but it is clear that the lowest-end models from each brand are priced that way for a reason.

Memory
  1. Kingston 0.20%
  2. Crucial 0.46%
  3. G.Skill 0.90%
  4. Corsair 1.08%
While the overall differences between brands aren't large, Corsair does make every single one of the modules with the highest return rates. The small effect on the overall return rate could mean that Corsair produces a lot of good products and a smaller number of very bad ones, compared to more consistent brands

Video Cards
  1. PNY 0.94%
  2. MSI 1.38%
  3. Gainward 1.61%
  4. Zotac 1.70%
  5. ASUS 1.81%
  6. Gigabyte 1.84%
  7. Sapphire 3.15%
The narrow differences between manufacturers that sell Geforce GTX cards is expected, as nVidia's "Project Greenlight" sets minimum standards that manufacturers must follow to ensure that their cards are reliable. For the last few issues Sapphire has been meaningfully worse than other manufacturers, so it's probably time to stop buying Sapphire products.

Hard Drives
Update:Backblaze pubished their numbers and found Seagate is poo poo.
  1. Seagate 0.95%
  2. Hitachi 1.16%
  3. Western Digital 1.19%
  4. Toshiba 1.54%
As we've known for awhile there is no statistically significant difference in reliability between brands. Seagate used to have much worse reliability due to firmware issues, but that doesn't seem to be apparent on any recent models. There is no reliability difference between larger and smaller capacities.

SSDs
  1. Samsung 0.28%
  2. Intel 0.63%
  3. Kingston 1.00%
  4. Corsair 1.88%
  5. Crucial 2.26%
  6. OCZ 2.27%
OCZ, Crucial, and Kingston are on the SSD Megathread's "Do Not Buy" list due to quality issues. OCZ has historically had the worst actual memory, some Kingston drives are also measurably lower quality in endurance test, but it seems that the failure rate of their HyperX 3K drives isn't reflected in overall return rates.

Alereon fucked around with this message at Jan 23, 2014 around 01:27

Adbot
ADBOT LOVES YOU

On3moresoul
Apr 22, 2010


This is probably a stupid question, I assume an exchange is the same as a return for this merchant's measurements?

Alereon
Feb 6, 2004

For me but LEFTHANDED

On3moresoul posted:

This is probably a stupid question, I assume an exchange is the same as a return for this merchant's measurements?
I don't know if it says specifically but yes, I'm quite sure it would be.

Aphrodite
Jun 27, 2006

Are you scared?

He's here.


drat, 12.67% for Sapphire's Radeon 7850.

iamnotcreative
Jul 28, 2002
What, you expected something creative here?

Aphrodite posted:

drat, 12.67% for Sapphire's Radeon 7850.

They were terrible. I had one for a while and it would lock my machine constantly; strangely it was usually desktop / web browsing / Youtube watching that did it. It apparently had something to do with the low clock they set for 2D stuff; if it changed suddenly it would lock the machine. I dealt with it for months and then ended up buying a 660 Ti; I really didn't want to dick with a return as it was a systemic problem with the 7850 line and not just my card.

beejay
Apr 7, 2002



Edit: Nevermind, there are numbers, and they are large. Carry on.

havenwaters
Nov 11, 2011

He has a ghost butt, for one thing, a GHOST BUTT, jade!

Put your hands down john


iamnotcreative posted:

They were terrible. I had one for a while and it would lock my machine constantly; strangely it was usually desktop / web browsing / Youtube watching that did it. It apparently had something to do with the low clock they set for 2D stuff; if it changed suddenly it would lock the machine. I dealt with it for months and then ended up buying a 660 Ti; I really didn't want to dick with a return as it was a systemic problem with the 7850 line and not just my card.

I guess I lucked out. My Sapphire 7850 didn't have those issues. It just had the good old 7xxx series texture flickering/corruption problem where black or white textures would flicker on the screen if your camera angle was in just the right position in DirectX 9 games. Drivers eventually fixed it in most games but some of them still mess up.

nftyw
Dec 27, 2006

It is a game... where you will put your life on the line.


I know I've seen a lot of bitching and complaining about Toshiba's hard drive quality and lack of customer support, I picked up an external USB3.0 750GB drive for $40 and people were complaining up and on about em lacking features and having basically no software, and looking at reviews for their desktop hard drives with their mediocre ratings averages make them seem like they are no better.

I'm honestly kind of surprised at the ~1% on Seagate considering all the moaning I hear about their drives, but when you have data loss, you really whine and moan about it (Western Digital drives failed on me often so I complained about theirs quite a bit). This computer originally had a Samsung that went bad, but I was able to get a Seagate shipped over from Best Buy, get most of my data pulled off and onto the new drive, and then returned the Samsung drive with basically no real problems or frustrations involved.

nftyw fucked around with this message at Nov 8, 2013 around 23:26

Setzer Gabbiani
Oct 13, 2004



Beyond the failures, Newegg and Sapphire making you deal with Athlon Micro for returns is probably a worse experience than the actual card dying itself

PainBreak
Jun 9, 2001


I had to return an MSI board for a really stupid reason. It was bundled with an FX8350, but came with a bios revision that didn't support the FX line. 1.2 supported it, but to go from 1.1 to 1.2, you needed an older processor.

My Googling led me to believe this isn't the first time this has happened, so I wonder what the return rate delta was on Intel vs AMD sockets. I see that, of the top 4 boards returned, 3 are AMD sockets.

Guni
Mar 11, 2010


PainBreak posted:

I had to return an MSI board for a really stupid reason. It was bundled with an FX8350, but came with a bios revision that didn't support the FX line. 1.2 supported it, but to go from 1.1 to 1.2, you needed an older processor.

My Googling led me to believe this isn't the first time this has happened, so I wonder what the return rate delta was on Intel vs AMD sockets. I see that, of the top 4 boards returned, 3 are AMD sockets.

Pretty sure that happened with the Z77 boards or some such had a problem like that, so it's definitely not just AMD that do it, to be fair.

Guni fucked around with this message at Nov 10, 2013 around 07:47

Panty Saluter
Jan 17, 2004

A Casual Gay


havenwaters posted:

I guess I lucked out. My Sapphire 7850 didn't have those issues. It just had the good old 7xxx series texture flickering/corruption problem where black or white textures would flicker on the screen if your camera angle was in just the right position in DirectX 9 games. Drivers eventually fixed it in most games but some of them still mess up.

Yeah me too. I've had a 7850 for four months now with no bad behavior that I've noticed. This is all pretty disappointing since I got a ton of mileage out of my old Sapphire 4850 (still running like a raped ape in the wife's computer) and the 7850 seemed like a pretty good card as well. Never dismiss dumb luck as a factor I guess.

Project Greenlight sounds like a pretty good idea and it seems to be bearing fruit. Makes a good case for jumping back to Nvidia with the next card I guess

Sh4
Feb 8, 2009


The merchant providing the stats is LDLC.com, largest IT web vendor in France. These numbers are reliable and based on high volumes.

HalloKitty
Sep 30, 2005

Adjust the bass and let the Alpine blast


Sh4 posted:

The merchant providing the stats is LDLC.com, largest IT web vendor in France. These numbers are reliable and based on high volumes.

Excellent.

Sapphire is kind of ruining the AMD card results all on its own.

Figures, really. Had a 7850 with awful coil whine.

An MSI 7850 didn't. Anecdotes, they may be, but as it turns out, there's more meat to it than that. Nice numbers.

vv Ahh, on that note, that card also made the microphone input incredibly noisy

HalloKitty fucked around with this message at Nov 11, 2013 around 15:24

Panty Saluter
Jan 17, 2004

A Casual Gay


HalloKitty posted:

Had a 7850 with awful coil whine.

I get a ton of crosstalk on my headphones ever since I've had the 7850. Of course that may be because the 7850 is so close to the sound card that they're almost touching.

Alereon
Feb 6, 2004

For me but LEFTHANDED

Backblaze posted an article on their experience with drive failures. They showed drive reliability starting to fall off after 3 years, where previous studies have shown the "elbow" to be at around the 5 year mark. It'll be interesting to see what their studies show as their drives age further. Additionally, one of their previous linked posts mentions that while the overall reliability seems to be the same, Seagate drives have worse behavior when they fail because they start causing issues before they start logging detectable errors. On the plus side, failing Seagate drives may keep limping longer than other drives.

Sh4
Feb 8, 2009


Alereon posted:

Backblaze posted an article on their experience with drive failures. They showed drive reliability starting to fall off after 3 years, where previous studies have shown the "elbow" to be at around the 5 year mark. It'll be interesting to see what their studies show as their drives age further. Additionally, one of their previous linked posts mentions that while the overall reliability seems to be the same, Seagate drives have worse behavior when they fail because they start causing issues before they start logging detectable errors. On the plus side, failing Seagate drives may keep limping longer than other drives.

That's pretty much what I observed too, I have a ghetto server with raid 10 of 500 Gb WD Caviar blue HD's, I always keep 5 drives active in the box because the failure rate is pretty high (daily time machine backups and bit torrent sync so the drives never even idle).

Anyway like half my HD's lasted less than a year but the ones that passed the 1 year mark made it at least 3 years.

Also these WD drives just start failing without any warning, I have weekly RAID checks and SMART alerts but all failures seems logic board based and they just stop working completely.

eXXon
Aug 19, 2002



Alereon posted:

Backblaze posted an article on their experience with drive failures. They showed drive reliability starting to fall off after 3 years, where previous studies have shown the "elbow" to be at around the 5 year mark. It'll be interesting to see what their studies show as their drives age further. Additionally, one of their previous linked posts mentions that while the overall reliability seems to be the same, Seagate drives have worse behavior when they fail because they start causing issues before they start logging detectable errors. On the plus side, failing Seagate drives may keep limping longer than other drives.

Those drives are probably being accessed quite regularly and powered on close to 24/7, no? So a typical home user who only keeps their computer on 6-12 hours per day, and say has an SSD and so keeps the HDDs spun down for more than half of that time... they're likely only racking up 3-4 hours of power on time per day. So if the failure rate while powered off is negligible, it seems reasonable to expect that your HDDs could last 6-10 years, not 2-3.

Cmdr Will Riker
Mar 27, 2003


Alereon posted:

ůSeagate drives have worse behavior when they fail because they start causing issues before they start logging detectable errors. On the plus side, failing Seagate drives may keep limping longer than other drives.

That's definitely been my anecdotal experience with Seagate drives recently. SMART passes without any issue while the drive starts locking up and doing all sorts of weirdness, but allowing me to backup pretty easily before RMAing.

Alereon
Feb 6, 2004

For me but LEFTHANDED

eXXon posted:

Those drives are probably being accessed quite regularly and powered on close to 24/7, no? So a typical home user who only keeps their computer on 6-12 hours per day, and say has an SSD and so keeps the HDDs spun down for more than half of that time... they're likely only racking up 3-4 hours of power on time per day. So if the failure rate while powered off is negligible, it seems reasonable to expect that your HDDs could last 6-10 years, not 2-3.
Google's 2007 study showed that utilization has no effect on drive lifespan, except that an already failing drive will finish dying more quickly under heavy load. Same deal with temperature, until you get to very high temps (>50C) it just doesn't make a difference.

eXXon
Aug 19, 2002



Alereon posted:

Google's 2007 study showed that utilization has no effect on drive lifespan, except that an already failing drive will finish dying more quickly under heavy load. Same deal with temperature, until you get to very high temps (>50C) it just doesn't make a difference.

Google study posted:

As is common in server-class deployments, the disks were powered on, spinning, and generally in service for essentially all of their recorded life.

I don't think that's the case for a typical home PC. With an SSD as an OS drive and power-saving features, HDDs can be off for a good chunk of the time a home PC is powered on, and I doubt they accrue much wear from sitting around powered off.

Alereon
Feb 6, 2004

For me but LEFTHANDED

eXXon posted:

I don't think that's the case for a typical home PC. With an SSD as an OS drive and power-saving features, HDDs can be off for a good chunk of the time a home PC is powered on, and I doubt they accrue much wear from sitting around powered off.
While it's logical to make that inference, I don't think that there's evidence that it's really true. A drive that isn't powering up or down is accruing fewer spinup/spindown cycles, which is a way that it's NOT wearing. It would likely be a big factor in poorly cooled external storage devices (which do exceed the safe operating ranges for drives), but probably not in computers. Also keep in mind that usage models that make it take longer to notice a failing drive don't mean that you're actually going longer between failures.

Alereon
Feb 6, 2004

For me but LEFTHANDED

Backblaze has published their harddrive reliability statistics: Seagate harddrives are poo poo, do not buy them. It's interesting that these results seem to be at odds with other failure rate statistics, but they published their numbers and they have enough drives for them to be statistically valid.

AlternateAccount
Apr 25, 2005
FYGM

Alereon posted:

Backblaze has published their harddrive reliability statistics: Seagate harddrives are poo poo, do not buy them. It's interesting that these results seem to be at odds with other failure rate statistics, but they published their numbers and they have enough drives for them to be statistically valid.

Holy poo poo, that's a pretty huge divergence. I sure hope their NAS drive line is more reliable :[

Geemer
Nov 4, 2010

What is it with Japan and ridiculous hair colors?


Alereon posted:

Backblaze has published their harddrive reliability statistics: Seagate harddrives are poo poo, do not buy them. It's interesting that these results seem to be at odds with other failure rate statistics, but they published their numbers and they have enough drives for them to be statistically valid.

Well gently caress, my 3TB drive has a 9.8% failure rate according to them. Let's hope that moderate desktop use that usually involves a lot of just being asleep and not really spinning most of the time raises the survivability.

Alereon
Feb 6, 2004

For me but LEFTHANDED

The one thing that would make me cautious about these numbers is BackBlaze's method for acquiring drives: They monitor for fatwallet and other deals with drives on deep discount, then crowdsource having people purchase the maximum and shipping them to Backblaze for a commission. I think more Seagate drives are more frequently on discount so they may be at the whims of individual drive shipping more than the other brands.

TWBalls
Apr 16, 2003
My medication never lies

AlternateAccount posted:

Holy poo poo, that's a pretty huge divergence. I sure hope their NAS drive line is more reliable :[

Pretty likely. They said the LP series did well. The Green and regular Barracuda series failed more often. Unsurprisingly, the greens failed the most. I'm actually surprised they had as "good" of luck with the 1.5TB drives, as I seem to recall Seagate had a pretty bad failure rate on those due to a lovely firmware. Although, I guess that might explain the high failure rate of the regular Barracudas?

Glad to heard HGST is still good. I have a 3TB drive I'm quite happy with. I actually bought it when I read an older article of theirs where they were recommending them.

Siochain
May 24, 2005

"can they get rid of any humans who are fans of shitheads like Kanye West, 50 Cent, or any other piece of crap "artist" who thinks they're all that?

And also get rid of anyone who has posted retarded shit on the internet."


TWBalls posted:

Pretty likely. They said the LP series did well. The Green and regular Barracuda series failed more often. Unsurprisingly, the greens failed the most. I'm actually surprised they had as "good" of luck with the 1.5TB drives, as I seem to recall Seagate had a pretty bad failure rate on those due to a lovely firmware. Although, I guess that might explain the high failure rate of the regular Barracudas?

Glad to heard HGST is still good. I have a 3TB drive I'm quite happy with. I actually bought it when I read an older article of theirs where they were recommending them.

The 7200.11 seagate firmware was primarily on the 500 and 750gb models made about 3-4 years ago. The shop I used to work at ended up RMA'ing approximately 150 out of 165 Seagate 7200.11 drives that we sold. It was brutal. Not a shock that Seagate sucks.

Agreed
Dec 30, 2003

The price of meat has just gone up, and your old lady has just gone down


I've got a Seagate 5900 RPM "green" 2TB that has been going strong since mid 2011 and shows no signs of errors or fault, which is an excellent counter as my sample size is n=1 and further, I

TWBalls
Apr 16, 2003
My medication never lies

They used to be my favorite brand, as I've had (and *knock on wood* continue to have) very few failures with them. We did have an issue a year or two ago with the Momentus Thins where, if the system was left on for ~10ish days, everything would slow to a crawl (30+ minutes to reboot. Didn't clear until the system was powered off. Then you had ~10 days again before it would come up). Here again, bad firmware.

I think I may stick with HGST for now.

GrizzlyCow
May 30, 2011


There are only Seagate, Toshiba, and Western Digital in the game now. You have to pick among them.

TWBalls
Apr 16, 2003
My medication never lies

GrizzlyCow posted:

There are only Seagate, Toshiba, and Western Digital in the game now. You have to pick among them.

Yep. Samsung are made by Seagate, HGST is owned by WD and Fujitsu sold their HDD division to Toshiba. Sad.

Agreed
Dec 30, 2003

The price of meat has just gone up, and your old lady has just gone down


When I finally get around to turning this box of parts into a computer, I'm getting WD Reds in at least 3TB capacity. Probably three of them. I've stayed away out of just not really understanding what the deal is with them but reliability seems great. I've used Caviar Black drives for years, some of my favorite drives ever were the 650ish GB WD Caviar Blacks from 2008. Things are still running without errors in a machine I have. Of course, I don't trust them, but, you know, they work fine, and for the time they were really fast. Single platter if I recall correctly.

HalloKitty
Sep 30, 2005

Adjust the bass and let the Alpine blast


TWBalls posted:

HGST is owned by WD

But, because of anti-monopoly legislation, some HGST drives are also made for Toshiba. You have to check individual models.

Wild EEPROM
Jul 29, 2011

i <3 lol cats


Alereon posted:

The one thing that would make me cautious about these numbers is BackBlaze's method for acquiring drives: They monitor for fatwallet and other deals with drives on deep discount, then crowdsource having people purchase the maximum and shipping them to Backblaze for a commission. I think more Seagate drives are more frequently on discount so they may be at the whims of individual drive shipping more than the other brands.

BackBlaze also buys drives that are inside an enclosure, and rip them out of the enclosures. You can only do this with Seagate and Hitachi drives; WD drives have the USB interface built right into the board, whereas the seagate and hitachi drives are bare drives with a sata to usb interface on them, and some seagate drives take it a step further and have a dock included, which you can use any sata drive with.
I'm guessing they're using a lot of those 3tb Seagate Expansion drives, the ones that are on sale every week for $99. Those seem to run hotter and louder than the 3tb desktop drives or 3tb backup plus drives, and only have a 1 year warranty instead of 2 years.

door.jar
Mar 17, 2010


Well, isn't this timely.

Alereon
Feb 6, 2004

For me but LEFTHANDED

door.jar posted:

Well, isn't this timely.

You're actually really lucky there, Seagate drives avoid reporting SMART errors to artificially lower their in-warranty failure rate, so usually the drive often has symptoms or dies before any logged errors are reported.

Tha Chodesweller
Jul 26, 2012

seriously dude chill out


Alereon posted:

You're actually really lucky there, Seagate drives avoid reporting SMART errors to artificially lower their in-warranty failure rate, so usually the drive often has symptoms or dies before any logged errors are reported.

I got lucky too then, I had a Seagate go out on me last year. Bought a WD Black to replace it and that's now throwing SMART warnings.

All hard drives are poo poo.

At least my Seagate was in warranty, and the Blacks have 5 year warranties.

Straker
Nov 10, 2005

professional sex-haver


door.jar posted:

Well, isn't this timely.

hey that's really pretty, what software is that?

I've seen (and probably have installed) all kinds of SMART monitoring crap but just never pay attention to any of it. It'd be nice to have something that will actually warn me coherently when anything important changes...

Adbot
ADBOT LOVES YOU

MrBond
Feb 19, 2004

FYI, Cheese NIPS are not the same as Cheez ITS

Alereon posted:

You're actually really lucky there, Seagate drives avoid reporting SMART errors to artificially lower their in-warranty failure rate, so usually the drive often has symptoms or dies before any logged errors are reported.

Hm is there a work around for this? I have a seagate laptop drive installed in my Mac and I've been suspicious of it lately.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply
«2 »