Please vote if you click on this

Silencing hard drives, optical drives and other storage devices

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply

What kind of permanent storage device do you have attached to your main computer, and has it failed within the last 12 months?

Poll ended at Fri Oct 08, 2010 8:14 pm

I have only an SSD, it has not failed
I have only an HDD, it has not failed
I have only an SSD, it has failed
No votes
I have only an HDD, it has failed
I have both an SSD and an HDD, only the SSD has failed
I have both an SSD and an HDD, only the HDD has failed
I have both an SSD and an HDD, neither have failed
I have both an SSD and an HDD, both have failed
Total votes: 35

Posts: 132
Joined: Sun Jan 31, 2010 6:43 pm
Location: US

Please vote if you click on this

Post by m1st » Wed Sep 08, 2010 8:14 pm

For any semblance of fair statistical representation, it is important that anybody who reads this votes. I'm trying to gauge SSD reliability relative to normal HDDs.

We all know about the 'Newegg effect', ie., a disproportionate amount of reviews are negative, since people are more likely to comment about something only if it broke. If everybody who reads this votes, then we should be able to more accurately measure actual reliability.

Since I'm trying to do this as a short poll, this will limit how precise this data actually is. As such, I have to limit the scope of the poll as follows:

When answering this poll:
1. Only answer regarding your main computer.
2. Only answer regarding your experiences in the last 12 months.

The reasons for these limitations are as follows:
1. Have you ever heard the phrase, "Everybody knows somebody who owns a Honda"? Although this is mostly true, the phrase, "Everybody owns a Honda" is definitely not true. The same could be true when talking about computer parts. The phrase, "Everybody has had a computer hardware failure" may be true, but it does not follow that every computer has a hardware failure (limiting the scope of the statement to normalized MTBF numbers). Lastly, if I asked the question, "Have you ever had an SSD failure", this might artificially inflate the SSD failure numbers. Therefore, we must only talk about HDDs/SSDs in your main computer system. This will limit, but not eliminate, the risks of inflated numbers. An addendum is below.

2. The reason for this limitation is threefold: One, I want data of more recent HDDs/SSDs. Two, I'm trying to find situations similar to my own hard drive mishaps. My SSDs have died within 6 months, so I'm fairly certain that this wasn't caused by the MLC NAND wearing out. I just want to see how many of these just die before their time. Lastly, since SSDs are much younger than HDDs, this may unfairly inflate the HDD failure rates, since failure rates seem to increase over time.

Final caveat: I'm using the logical (as in Philosophy logic class) definition of the word 'some', id est, "At least one". Therefore, if I say, "I've had some SSDs fail", having only one SSD fail makes that statement true.

Getting to the end of what was supposed to be a short poll, I want to talk a little bit about the hopeful scope of this poll. This will only poll failure rates from the last 12 months, and will only poll failure rates on our main computers. One can see that there is an imprecision in this poll - it does not account for those of use who have multiple SSDs/HDDs in our computers. If I had 16 SSDs in my computer and one HDD, then my chances of reporting an SSD failure would be much higher than someone who had just one SSD in their computer. I realize this is an issue, but this is a tradeoff between poll precision and sample size. I'm already eliminating a large potential sample pool by limiting the scope of the poll to only our main computers. I don't think the added precision of talking about individual hard drives in our computers would offset the loss of sample size

Definition: SSD market penetration is going to be drastically lower than HDD market penetration. I will normalize the failure rates of each technology to the replies regarding each penetration. So if I only end up with 2 SSD failures and 8 HDD failures, this doesn't mean that SSDs are inherently more reliable. We will have to see how big the polling size is for each technology before we determine anything.


So finally:

What kind of permanent storage device do you have attached to your main computer, and has it failed within the last 12 months?

Please feel free to elaborate about the nature of your computers below. I know I'll be posting my sob story :-P

Posts: 132
Joined: Sun Jan 31, 2010 6:43 pm
Location: US

Post by m1st » Wed Sep 08, 2010 8:26 pm

And my story...

In my main computer, I have two drives:

640GB Western Digital Caviar Blue (WD6400AAKS)
120GB OCZ Agility 2 (Sandforce)

My WD is older than my OCZ and was originally my system drive. I then bought the OCZ and converted the WD into a data drive. My OCZ failed within 2 days of first using it. I returned it and got a new one. This one worked for 3 months, but has now just died. Both SSDs exhibited typical drive warning signs (programs began to lag and then start responding again). Then my computer blue-screened and shut down. When I rebooted, the drive was no longer visible during POST. I unplugged it, used different cables, and watched some tennis, but none of these troubleshooting steps allowed the drive to be detected at POST. I left it unplugged for longer and it was visible during POST, but then it crashed again upon bootup and was no longer visible. I have backups so that's not the problem, but it is annoying that I'm without a computer for the forseeable future...

As an aside, even with two failed SSDs in the last 3 months, the speed increase is AMAZING. True, it's only really noticable during bootup and when opening a lot of programs at the same time since Windows Vista/7 caches everything anyways, but it's really nice to not dread rebooting your computer. And my computer isn't a slouch either.

And in an enterprise environment, SSDs are just worth it. As long as they don't crash every 3 months...

Friend of SPCR
Posts: 1346
Joined: Tue Jul 08, 2008 4:53 am
Location: CT

Post by frenchie » Thu Sep 09, 2010 2:43 am

80GB intel G2 SSD as OS drive and Samsung F1 1TB as main storage. Neither as failed.

Friend of SPCR
Posts: 378
Joined: Mon Dec 12, 2005 3:29 am
Location: Melbourne, Australia

Post by Eunos » Thu Sep 09, 2010 4:50 am

Sorry to see you've had a bad run with the OCZ Sandforce drive. Overall they are regarded very reasonably, and in terms of performance/$ they are top class. I'd probably buy one myself, but my X25-M just won't die. :twisted:

My only constructive criticism of the poll is that the model of hard drive or SSD may be a greater factor than whether it uses platters or flash cells. There have been legends and lemons from both camps.

Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Please vote if you click on this

Post by quest_for_silence » Thu Sep 09, 2010 10:17 am

m1st wrote:For any semblance of fair statistical representation, it is important that anybody who reads this votes. I'm trying to gauge SSD reliability relative to normal HDDs.

I guess you should modify the topic and give more self explanatory subject to your poll/thread.


*Lifetime Patron*
Posts: 336
Joined: Thu Nov 09, 2006 4:44 pm
Location: San Jose, CA

Post by kittle » Thu Sep 09, 2010 12:00 pm

I recently got a new PC and jumped on the SSD bandwagon. Intel X25 80GB and a WD cavair black 2TB.
About 2months after I got the PC, the WD drive started dropping offline for no visible reason. all diagnostics I ran came back with no errors, so I RMA'd it and its been fine so far.

Now contrast that to my older desktop/server system which has 6 10k rpm SCSI drives of various sizes from WD, Seagate and Maxtor.
the first HD I got back in 2001, failed 6months after the 5yr warranty expired. the 2nd HD i got in 2002 is still running 8 years later. I got another batch of HDs in 2005 when I got a new system. All of them are still running fine after 5yrs of almost 24/7 operation.

Based on my experience, the older, slower SCSI drives have a MUCH better track record than the newere SATA drives

for SSDs -- with a sample size of 1, I cant say much other than its VERY fast.

Posts: 132
Joined: Sun Jan 31, 2010 6:43 pm
Location: US

Post by m1st » Fri Sep 10, 2010 5:30 pm

Thanks to everybody for the votes/feedback so far.

Polling by manufacturer/controller would most definitely be extremely enlightening, but I was more interested at SSDs vs HDDs as a whole. It would be interesting to see which SSD controller seems to have a higher reliability rate, however...

I thought that making a more explanatory title would bias the results, but upon thinking about it more closely, you're right. For next time :-P

Friend of SPCR
Posts: 279
Joined: Sun May 06, 2007 10:37 am

Post by PartEleven » Fri Sep 10, 2010 7:16 pm

For SSDs specifically, I think it may be more prudent to poll by controller type rather than manufacturer. Failure typically lies with the controller software rather than poor manufacturing. What complicates things is that sometimes manufacturers will tweak the controller themselves or use a model on special order from the controller producer.

Posts: 891
Joined: Thu Sep 07, 2006 11:20 pm
Location: Poland

Post by kater » Fri Sep 10, 2010 11:15 pm

Never had and SSD so far, but had plenty of HDDs. For daily work on my main PC, for my HTPC, for archiving, storage and moving data. Last time a HDD died on me (nearly causing a cardiac arrest and effectively cutting my time on Earth by at least 3 years...) was about 8 years ago. It was an IBM Deathstar. Granted, few (2?) HDDs started showing isolated bad sectors but never died.

Post Reply