It's hard to say. For the original Vertex 3, OCZ reports Iometer and AS-SSD numbers, but for the new Vertex 3.20 OCZ only reports one set of unspecified numbers. I'm assuming they are Iometer scores and hence I'm using Iometer scores for the Vertex 3. Anyway, it could be a difference in firmware too.
I currently have quite a few drives to review so I'm not sure if I'll have time to squeeze another drive in. I'll try my best though :-)
I don't know why I took the time out of my work day to write this on the internet in a comment section.. To make it look right I even had to learn how to insert a "tab" space in a text area without the browser selecting the next element on the page. (Keyboard shortcut Alt + 09). Vertex 3.20 Vertex 3 Capacity 120GB 240GB 120GB 240GB 4KB Random Read 20K IOPS 35K IOPS 20K IOPS 40K IOPS 4KB Random Write 40K IOPS 65K IOPS 60K IOPS 60K IOPS
Anyway long story short: the 40k should've said 20k. The rest was correct.
Apparently the comment system removed my painfully organized tab-table.. It also thought my comment was spam, so I had to remove the source links to make it pass.
According to OCZ's site, that's what the specifications are. SandForce has always been a bit weird, for example 480GB Vertex 3 is noticeably slower than the 240GB model, even though more NAND should result in increased parallelism and hence enhanced performance.
Since our comment system thinks every link is spam, I can't link the datasheet but just check the Vertex 3 product brief PDF at OCZ's site.
"unlike with the Vertex 2 when OCZ silently switched to 25nm NAND" But they didn't just switch to 25nm NAND. They took out some speed AND capacity. They should have done a mass recall but they didn't.
Only people that knew they were scammed and complained managed to get their issue resolved.
"But they didn't just switch to 25nm NAND. They took out some speed AND capacity. They should have done a mass recall but they didn't." A side effect of the switch. "Only people that knew they were scammed and complained managed to get their issue resolved." One could say that those other people were fine with the product they got. Or they made the decision that their time and effort was better spent on other problems they might have.
3Xnm die was 4GB, whereas 25nm IMFT NAND moved to 8GB die. That resulted in decreased performance in smaller capacity drives because the increase in die size led to less parallelism.
It was definitely wrong to do the switch silently because it had an enormous impact on performance. The good news is that OCZ isn't doing it again.
I know the technical details. The effect of their actions is the most important thing and it wasn't just performance, capacity was affected as well (they would have picked that up in the smoke testing stage).
The bad news is that they never did a mass recall so no one other, than OCZ, know how many affected drives are out there in circulation.
The 25nm Vertex 2 drives had less capacity than the original V2s but the SKUs were the same. Can you spot what OCZ did wrong?
"The matter is complicated by the way SandForce's NAND redundancy works. The SF-1000 series controllers have a feature called RAISE that allows your drive to keep working even if a single NAND die fails. The controller accomplishes this redundancy by writing parity data across all NAND devices in the SSD. Should one die fail, the lost data is reconstructed from the remaining data + parity and mapped to a new location in NAND. As a result, total drive capacity is reduced by the size of a single NAND die. With twice the density per NAND die in these early 25nm drives, usable capacity was also reduced when OCZ made the switch with Vertex 2."
All I hear are horror stories about OCZ SSDs and how frequently they fail. They have gotten really affordable but I would rather pay an extra $10-20 for brands that are more proven in their reliability. Hopefully they turn things around over there.
That is an over-simplification. I have only ever heard about the cheap OCZ drives like the Agility failing, so I think their restructuring is long overdue. The Vertex line is very reliable.
I have personally installed 6 OCZ vertex drives into my systems and never had a *significant* issue. Most reports just speak about failed OCZ drives, but fail to list specific models. I have one guy that bought an OCZ Solid that failed. RMA'd and got an Agility, it failed. RMA'd and got another Agility, it failed. I told him to give up on the warranty and just by a drive with a better track record like Intel 335.
While I use OCZ Vertex and Vectors and like them, I still feel that Intel drives are more reliable. At least more consumer friendly. I have built every computer for the last year with Intel SSD's and never got a blip from them. Well, once, but that turned out to be an OS install issue that was fixed by just reinstalling Windows. My personal systems are all getting Vector drives currently, I just wish OCZ had a Windows SSD Tools program like the Intel SSD Toolbox. The OCZ ISO is aggravating, it is alien in so many ways.
OCZ core series 30GB (~9 month in my desktop, used instead of an USB-stick ever since) OCZ core series V2 60GB (1 year in my desktop, then 1 year in my notebook, now sitting in my bedroom-mediaplayer) OCZ Vertex 2 120GB (1 year in my desktop, 1 year in my notebook, no in my fileserver for close to 1 year) OCZ Vertex 3 120GB (1 year in my desktop, now in my notebook for a few month)
Never had any of those drives outright fail or loose data. The core series drives got slowed down with time, and Win XP did occasionally crash when it had to wait too long for a response, but that was pretty much standard for MLC drives back then.
I did now switch to a Samsung 840Pro 240GB now for my desktop, but thats because it promised performance beyond anything OCZ offers right now, not because I was unhappy with OCZ.
There. Now you can no longer say that all you hear are horror stories.
The problem wasn't OCZ exactly. reliability issue was largely due to failure of that Sandforce controller.
CZ was the most aggressive manufacturer in terms of introducing drives using the SF controller. This gave them market dominance, so when SF-based drives started falling in large numbers the majority happened to have OCZ branding.
Now, OCZ customer service horror stories, that is all on them. The Vertex 2 bait-and-switch was the most egregious but was worsened by their fumbling the response until mass customer fury forced a full-fledged retreat.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
26 Comments
Back to Article
mayankleoboy1 - Wednesday, February 20, 2013 - link
How come some of the values have improved for 120GB ? Firmware magic, or moar over provisioning, or a combination of the two ?Will you do a re-review ?
Kristian Vättö - Wednesday, February 20, 2013 - link
It's hard to say. For the original Vertex 3, OCZ reports Iometer and AS-SSD numbers, but for the new Vertex 3.20 OCZ only reports one set of unspecified numbers. I'm assuming they are Iometer scores and hence I'm using Iometer scores for the Vertex 3. Anyway, it could be a difference in firmware too.I currently have quite a few drives to review so I'm not sure if I'll have time to squeeze another drive in. I'll try my best though :-)
ahar - Wednesday, February 20, 2013 - link
Do you have time to correct "less products" to "fewer products"? ;-)Kristian Vättö - Wednesday, February 20, 2013 - link
Sure, thanks for the heads up! :-)ShieTar - Wednesday, February 20, 2013 - link
Is it correct that the 240GB version has lower random read speeds (35k) than the 120GB version (40k)? It would seem more logical the other way around.Marlowe - Wednesday, February 20, 2013 - link
I don't know why I took the time out of my work day to write this on the internet in a comment section.. To make it look right I even had to learn how to insert a "tab" space in a text area without the browser selecting the next element on the page. (Keyboard shortcut Alt + 09).Vertex 3.20 Vertex 3
Capacity 120GB 240GB 120GB 240GB
4KB Random Read 20K IOPS 35K IOPS 20K IOPS 40K IOPS
4KB Random Write 40K IOPS 65K IOPS 60K IOPS 60K IOPS
Anyway long story short: the 40k should've said 20k. The rest was correct.
Marlowe - Wednesday, February 20, 2013 - link
Apparently the comment system removed my painfully organized tab-table.. It also thought my comment was spam, so I had to remove the source links to make it pass.Kristian Vättö - Wednesday, February 20, 2013 - link
According to OCZ's site, that's what the specifications are. SandForce has always been a bit weird, for example 480GB Vertex 3 is noticeably slower than the 240GB model, even though more NAND should result in increased parallelism and hence enhanced performance.Since our comment system thinks every link is spam, I can't link the datasheet but just check the Vertex 3 product brief PDF at OCZ's site.
Kristian Vättö - Wednesday, February 20, 2013 - link
Oops, my bad, you're correct. The 120GB is rated at 20K IOPS for random read, not 40K. Fixed.Marlowe - Wednesday, February 20, 2013 - link
Well all links are blocked except those in the "Love my job"-posts ;-)semo - Wednesday, February 20, 2013 - link
"unlike with the Vertex 2 when OCZ silently switched to 25nm NAND"But they didn't just switch to 25nm NAND. They took out some speed AND capacity. They should have done a mass recall but they didn't.
Only people that knew they were scammed and complained managed to get their issue resolved.
Death666Angel - Wednesday, February 20, 2013 - link
"But they didn't just switch to 25nm NAND. They took out some speed AND capacity. They should have done a mass recall but they didn't."A side effect of the switch.
"Only people that knew they were scammed and complained managed to get their issue resolved."
One could say that those other people were fine with the product they got. Or they made the decision that their time and effort was better spent on other problems they might have.
semo - Wednesday, February 20, 2013 - link
Basically you are saying that if OCZ managed to get away with it, more power to them?Kristian Vättö - Wednesday, February 20, 2013 - link
3Xnm die was 4GB, whereas 25nm IMFT NAND moved to 8GB die. That resulted in decreased performance in smaller capacity drives because the increase in die size led to less parallelism.It was definitely wrong to do the switch silently because it had an enormous impact on performance. The good news is that OCZ isn't doing it again.
semo - Wednesday, February 20, 2013 - link
I know the technical details. The effect of their actions is the most important thing and it wasn't just performance, capacity was affected as well (they would have picked that up in the smoke testing stage).The bad news is that they never did a mass recall so no one other, than OCZ, know how many affected drives are out there in circulation.
Kristian Vättö - Wednesday, February 20, 2013 - link
What do you mean by capacity?semo - Thursday, February 21, 2013 - link
The 25nm Vertex 2 drives had less capacity than the original V2s but the SKUs were the same. Can you spot what OCZ did wrong?"The matter is complicated by the way SandForce's NAND redundancy works. The SF-1000 series controllers have a feature called RAISE that allows your drive to keep working even if a single NAND die fails. The controller accomplishes this redundancy by writing parity data across all NAND devices in the SSD. Should one die fail, the lost data is reconstructed from the remaining data + parity and mapped to a new location in NAND. As a result, total drive capacity is reduced by the size of a single NAND die. With twice the density per NAND die in these early 25nm drives, usable capacity was also reduced when OCZ made the switch with Vertex 2."
http://www.anandtech.com/show/4256/the-ocz-vertex-...
http://forums.anandtech.com/showthread.php?t=21433...
HisDivineOrder - Wednesday, February 20, 2013 - link
...ran out of 3. Plus, Sandforce is probably pretty easy to get hold of now.smitty123 - Wednesday, February 20, 2013 - link
... with the latest vertex 4 and vector firmwaresand specify what firmware version of the other vertex 4's used in the bench,
there's a vertex4 v1.4 and another.
Hrel - Wednesday, February 20, 2013 - link
So, is the price lower now? I want a 512GB SSD for under 200 bucks!Paulman - Wednesday, February 20, 2013 - link
Yes - I, too, am very curious about what sort of cost savings to expect (and how soon) :POwls - Wednesday, February 20, 2013 - link
All I hear are horror stories about OCZ SSDs and how frequently they fail. They have gotten really affordable but I would rather pay an extra $10-20 for brands that are more proven in their reliability. Hopefully they turn things around over there.'nar - Thursday, February 21, 2013 - link
That is an over-simplification. I have only ever heard about the cheap OCZ drives like the Agility failing, so I think their restructuring is long overdue. The Vertex line is very reliable.I have personally installed 6 OCZ vertex drives into my systems and never had a *significant* issue. Most reports just speak about failed OCZ drives, but fail to list specific models. I have one guy that bought an OCZ Solid that failed. RMA'd and got an Agility, it failed. RMA'd and got another Agility, it failed. I told him to give up on the warranty and just by a drive with a better track record like Intel 335.
While I use OCZ Vertex and Vectors and like them, I still feel that Intel drives are more reliable. At least more consumer friendly. I have built every computer for the last year with Intel SSD's and never got a blip from them. Well, once, but that turned out to be an OS install issue that was fixed by just reinstalling Windows. My personal systems are all getting Vector drives currently, I just wish OCZ had a Windows SSD Tools program like the Intel SSD Toolbox. The OCZ ISO is aggravating, it is alien in so many ways.
ShieTar - Thursday, February 21, 2013 - link
Well, personally I have bought four SSDs so far:OCZ core series 30GB (~9 month in my desktop, used instead of an USB-stick ever since)
OCZ core series V2 60GB (1 year in my desktop, then 1 year in my notebook, now sitting in my bedroom-mediaplayer)
OCZ Vertex 2 120GB (1 year in my desktop, 1 year in my notebook, no in my fileserver for close to 1 year)
OCZ Vertex 3 120GB (1 year in my desktop, now in my notebook for a few month)
Never had any of those drives outright fail or loose data. The core series drives got slowed down with time, and Win XP did occasionally crash when it had to wait too long for a response, but that was pretty much standard for MLC drives back then.
I did now switch to a Samsung 840Pro 240GB now for my desktop, but thats because it promised performance beyond anything OCZ offers right now, not because I was unhappy with OCZ.
There. Now you can no longer say that all you hear are horror stories.
Galcobar - Sunday, February 24, 2013 - link
The problem wasn't OCZ exactly. reliability issue was largely due to failure of that Sandforce controller.CZ was the most aggressive manufacturer in terms of introducing drives using the SF controller. This gave them market dominance, so when SF-based drives started falling in large numbers the majority happened to have OCZ branding.
Now, OCZ customer service horror stories, that is all on them. The Vertex 2 bait-and-switch was the most egregious but was worsened by their fumbling the response until mass customer fury forced a full-fledged retreat.
Tjalve - Tuesday, February 26, 2013 - link
Great... 2 new models for the Vertex 3... How many is there now?Vertex 3 60GB
Vertex 3 LP 60GB
vertex 3 90GB
Vertex 3 120GB
Vertex 3 LP 120GB
Vertex 3 MAX IOPS 120GB
Vertex 3.20 120GB
Vertex 3 128GB
Vertex 3 240GB
Vertex 3 LP 240GB
Vertex 3 MAX IOPS 240GB
Vertex 3.20 240GB
Vertex 3 256GB
Vertex 3 480GB
Vertex 3 LP 480GB
Vertex 3 512GB
or did i miss someone? :P