Is it OK to buy OCZ SSDs now?

blckgrffn

Diamond Member
May 1, 2003
9,517
3,881
136
www.teamjuchems.com
http://www.anandtech.com/show/5741/ocz-confirms-octane-and-vertex-4-use-marvell-based-silicon

I have been generally happy with the 128GB Petrol drive I bought from MC for ~$100 AMIR, and was thinking about buying another. My previous rule of thumb was "no sandforce" so I think this still works?

FWIW, http://alex-is.de/PHP/fusion/downloads.php?download_id=9 causes my Petrol drive to become unresponsive, it doesn't finish. That does bother me. Latest firmware as of ~6 weeks ago.

Any Petrol/Octane lovers/haters out there? Or can OCZ drives still DIAF?
 

joshhedge

Senior member
Nov 19, 2011
601
0
0
No problems with my 120GB OCZ Agility 3... Other than for some reason it hit 120 degrees Celsius at one point... no idea why. My other 120GB Corsair Force 3 gets better reads and writes though.
 

mikeymikec

Lifer
May 19, 2011
19,439
13,001
136
@ joshhedge

Where did you get that reading from? I'm wondering about the possibility of an incorrect reading.
 

joshhedge

Senior member
Nov 19, 2011
601
0
0
Haha, I was a tad bit concerned! I was actually doing a disk bench to check I was getting decent reads/writes after I installed it and a warning popped up on Win7s desktop. That was around 3 months ago and its never happened since.
 

groberts101

Golden Member
Mar 17, 2011
1,390
0
0
There's no temperature sensor in the drive and it's just a static number without any meaning.

If you're concerned about it or like to run benchmarks designed for HDD that would report it?.. go get the Linux update tool to fix it. Then apply the included temp fix which will set it to another static number that won't freak you out.

About the only time I've ever seen issues result from it would be on some raidcards and server boards smart based utilities though.
 

blckgrffn

Diamond Member
May 1, 2003
9,517
3,881
136
www.teamjuchems.com
I noticed the petrol doesn't report a bunch of values that my other ssds (samsung 470, Intel G2, Crucial M4) do. What's up with that? Laziness on OCZ's part? Is this consistent across their offerings?
 
Last edited:

joshhedge

Senior member
Nov 19, 2011
601
0
0
There's no temperature sensor in the drive and it's just a static number without any meaning.

If you're concerned about it or like to run benchmarks designed for HDD that would report it?.. go get the Linux update tool to fix it. Then apply the included temp fix which will set it to another static number that won't freak you out.

About the only time I've ever seen issues result from it would be on some raidcards and server boards smart based utilities though.

Thanks for the advice, I'm not bothered about it now, if it occurs again I'll follow that fix :)
 

alexruiz

Platinum Member
Sep 21, 2001
2,836
556
126
http://www.anandtech.com/show/5741/ocz-confirms-octane-and-vertex-4-use-marvell-based-silicon

I have been generally happy with the 128GB Petrol drive I bought from MC for ~$100 AMIR, and was thinking about buying another. My previous rule of thumb was "no sandforce" so I think this still works?

FWIW, http://alex-is.de/PHP/fusion/downloads.php?download_id=9 causes my Petrol drive to become unresponsive, it doesn't finish. That does bother me. Latest firmware as of ~6 weeks ago.

Any Petrol/Octane lovers/haters out there? Or can OCZ drives still DIAF?

I got 2 of the same SSD (Petrol 128GB) just a few weeks ago based on the same premise of "no sandforce". Let me tell you my experience.

My first SSD was a Kingston SSDnow SNv425 64GB (jmicrom JM112?) already considered bottom of the barrel of the gen 1st SSDs. It was however night and day compared to the WD scorpio black I had before in the laptop, and the SSD was rock solid. Eventually I upgraded to a Kingston hyper-X 120GB (SF-2281 syncronous 25 nm) considered by several sites as one of the top current generation SSDs. The upgrade was driven by size rather than speed, and in all honesty, the only difference in speed could be seen in benchmarks and WEI because the machine felt exactly the same during daily usage.

The hyper-x, however, had this weird random "hang" once in a while, where the system would hang for 4-5 secs before responding again. Hangs would be occasional (once every 2 weeks or so) and because of that I sold the drive and went Indilinx with the Petrol. Reliability over speed was key, and I got them despite the dreadful reviews at the egg. I barely give weight to those reviews as it is very common to see 1 egg reviews because "windows didn't see the drive, never mounted a letter..." (how about partitioning, formatting and assigning the partition first?)

One of the Petrols is in my laptop, the other in my desktop. Firmware got upgraded right after opening the drives. Right off the bat the one for my desktop was already reporting "caution" health status from within crystaldiskinfo. Firmware upgrade obviously cleared everything, but the firmware reflash took 2 tries in that same SSD to be successful. The other one was perfect, so I popped that one in my laptop.

The problematic one in my desktop loses the BCD info for win 7 twice per week, asking for a windows repair. I have only programs, and my backup image is up to date so that is why I risked it as the image restore takes only a few minutes. Still, it was already annoying to do the windows repair twice a week. It benched decently fast, and the WEI was 7.7 in a SATA 3 port off the SB950.

The other one in my laptop was a stark contrast, always smooth, always flawless. I have some important data on it, and while my backups are up to date, still, a week lost of data would be a week lost. The SSD, however, would not give me reason to worry so I was getting complacent. Then, one day, out of the blue, when starting the machine, I get a "chkdsk is checking drive d: for consistency". I made a backup of the data right after that. I get the annoying chkdsk once a week now. Twice a week backups... the SSD is already dying.

If your drive is working fine, just wait. It seems the NAND used in the petrol was the cheapest OCZ could get. I don't know if the nand is the same in the agility 3, or if other manufacturers use the same nand and will get hit also. Firmware seems fine, so I would say, yes, you can buy a OCZ SSD safely, just make sure is a vertex ;)
 

thelastjuju

Senior member
Nov 6, 2011
444
2
0
I just don't understand why anybody would take the RISK.

Are they discounting and rebating them that heavily or something?

We have guys like Intel, Samsung, Crucial, and Corsair, with a consistent record of manufacturing reliable SSD's, and then we have OCZ.. with the most alarming failure rate among SSD technology.

:\
 

groberts101

Golden Member
Mar 17, 2011
1,390
0
0
They are just as reliable now as they were years ago ;P

That's very true and I couldn't have said it better myself.

My 15 older drives are functioning just as fast(actually even faster due to free firmware speed upgrades) and haven't given me any issues whatsoever. ():)
 

chloros

Member
Feb 1, 2011
95
0
0
I just don't understand why anybody would take the RISK.

Are they discounting and rebating them that heavily or something?

We have guys like Intel, Samsung, Crucial, and Corsair, with a consistent record of manufacturing reliable SSD's, and then we have OCZ.. with the most alarming failure rate among SSD technology.

:\

Im with you on this. Ive only gone intel and have built many machines for friends. Never one issue. Why deal with the risk when its right in your face?
 

kmmatney

Diamond Member
Jun 19, 2000
4,363
1
81
I haven't had any probelms with several Agility 2 drives I purchased over a year ago. However, you can get Samsung or Crucial drives for only a little bit more, so why risk it one a cheap brand.
 

Railgun

Golden Member
Mar 27, 2010
1,289
2
81
Everyone reads up on issues from those that are more vocal about their failures, and without a lot of first hand experience, run for the hills once OCZ is brought up.

STILL running strong with 4x in a RAID (over a year) and 1 in a laptop (about six months). All three are Vertex 3 120G variants.

The fact of the matter is a failure without any description to the HW (controller) used, drive usage (capacity utilized) and general usage, it somewhat useless. For all we know, the drive was full to capacity. Was not up to date with its FW. Ran 200 benchmark tests. Was running while a power outage occured. There are many variables.

As I mentioned above, I'm using what I would consider a proper controller, an Areca 1880i. Not some on board controller where I've seen the majority of failures occur. But again, it's not the full story.

Personally, I'd say you're fine with OCZ. But that's my experience anyway.
 

Cerb

Elite Member
Aug 26, 2000
17,484
33
86
Everyone reads up on issues from those that are more vocal about their failures, and without a lot of first hand experience, run for the hills once OCZ is brought up.

STILL running strong with 4x in a RAID (over a year) and 1 in a laptop (about six months). All three are Vertex 3 120G variants.

The fact of the matter is a failure without any description to the HW (controller) used, drive usage (capacity utilized) and general usage, it somewhat useless.
Unless a specific bug with a specific controller is found (such as M4's panic with some AMD SBs), or unless the controllr is known to have unfixed bugs, the controller making a difference is the drive's fault. In that case, I would only give them a pass if the controller came out after the drive did, in which case a preventative fix would have been impossible (which, in Crucial's case, did happen).
For all we know, the drive was full to capacity.
At which point the OS will complain, the user can delete files, and nothing should go wrong with the drive itself. If it does, the drive is defective.
Was not up to date with its FW.
Given that most FW updates include bug fixes, this one would be legitimate.
Ran 200 benchmark tests.
The user should be able to run benchmark tests until the drive decides it must go read-only. If the drive can't handle that, it is defective.
Was running while a power outage occured.
Unless it got hit by a power surge, in which case it's probably not the only fried part, that should not cause a failure for the drive. If it can't recover from a power outage while writing, the drive is defective.
There are many variables.
Most of which should have been caught and fixed before consumers got access to the drives. All that extra info may be necessary for filing a good bug report, but it is not necessary to report a failure. Trying to access the drive from another computer after an apparent failure, to confirm that it's broken; and/or trying to wipe it (possibly including a fresh FW flash) and use it again, to confirm it can still function, are the most you may need for a drive failure.
 

Railgun

Golden Member
Mar 27, 2010
1,289
2
81
The OS complaining is irrelevant as we don't know if that was the case or not.

Would 2000 benchmark tests be better? Back to back, continuously?

You can guarantee, 100%, that a drive will be OK 100% of the time after 100% of power failure tests? Regardless of whether it's a HDD or SSD.

The point is there's no definitive metric to any particular failure rate. Could simply be MTBF being hit. How many drives actually exist in consumer's hands? How many on current FW are an issue? Are they SF 1200 or 2200 controllers? Do we actually have a metric on that?

Yeah, it's not necessary to report a failure, but as we're an over analyzing bunch, how are you going to make a determination that OCZ sucks simply based on incomplete data? BMW just recalled 1.3 million cars. Doesn't mean no one should buy one ever again.

Firestone had a huge problem with their tires, yet Bridgestone is still a very popular tire manufacturer and OEM supplier.

And again, there are countless others that are running these drives that haven't had any issues. So, while I appreciate those that have apprehension towards OCZ, the fact that they've failed for x reason shouldn't be justification enough, especially for those that haven't had first hand experience with it and without knowing circuimstances surrounding the failures.
 

blckgrffn

Diamond Member
May 1, 2003
9,517
3,881
136
www.teamjuchems.com
I got 2 of the same SSD (Petrol 128GB) just a few weeks ago based on the same premise of "no sandforce". Let me tell you my experience.

*truncated*

If your drive is working fine, just wait. It seems the NAND used in the petrol was the cheapest OCZ could get. I don't know if the nand is the same in the agility 3, or if other manufacturers use the same nand and will get hit also. Firmware seems fine, so I would say, yes, you can buy a OCZ SSD safely, just make sure is a vertex ;)

Thank you for sharing! It looks like I'll be able to stave off the urge to buy another... I am not sure what to do with the one I have. It is being used right now in my ESXi "teaching" box so that ~10 students can build VMs at once on it, it seems to work well for that (compared to a spinning drive - wow, night and day) but now I am wondering how to use it in the long term...

Again, thank you for your input - as well as everyone else who has chimed in.
 

Minerva

Platinum Member
Nov 18, 1999
2,127
19
81
It's not OK to buy ANYTHING with the three letters (OCZ) IMHO. Horrible company, avoid like the plague x 1000...

Now that I've put away my soapbox let me tell you this.

I don't have a problem with the Indilinx controller. Matter of fact I have a bunch of Gskill Falcons (remember those from 2009?) Indilinx Barefoot drives. They have been going strong the entire time on both desktops and laptops. No complaints. Been through all the firmware updates.

Can't say I've been pleased with SF disks, at all.

Probably OK with the Vertex 4 BUT you do take your risks if you need support from them. Hardware-wise it SHOULD be solid but they are so new it's hard to say. All drives have had their fair share of problems. Look at the Crucial M4. It seems like it's quite mature with the latest update. If you don't have 0309 and the drive has 5000 hours on its clock it disappears on the hour. Now that is a ticking time bomb. But at least the fix is easy...
 

alexruiz

Platinum Member
Sep 21, 2001
2,836
556
126
Thank you for sharing! It looks like I'll be able to stave off the urge to buy another... I am not sure what to do with the one I have. It is being used right now in my ESXi "teaching" box so that ~10 students can build VMs at once on it, it seems to work well for that (compared to a spinning drive - wow, night and day) but now I am wondering how to use it in the long term...

Again, thank you for your input - as well as everyone else who has chimed in.

You are very welcome!

I thought the SSD in my laptop was one of the lucky ones. It just took a little longer to show up the issues. It doesn't mean, however, that yours will do the same. There are a lot of factors that we don't know, and something as simple as a batch of bad NAND could be the culprit. Your drive might end up being fine

If I was buying another one right now, I would just be on the hunt for a Vertex 3, or any other sync / toggle NAND SSD. Wait, I just did that, I have 2 vertex 3 120GB en route, as well as 2 sandisk extreme 120GB. I'll RMA the petrols once I make the switch.


Alex

Edit: Look what popped in my email inbox :)
http://www.newegg.com/Product/Product.aspx?Item=N82E16820233191
 
Last edited: