OCZ Shipping 25nm SSDs, customers not happy over lack of change?

Came across this article over @ storagereview. Looks interesting. This is what caught my eye...

This is obviously an issue. I'm a bit amazed OCZ went this route.

full article is here -
SSDs Shifting to 25nm NAND - What You Need to Know | StorageReview.com

Thoughts?

 

I am surprised the article said that the cost reduction is only that small(< 20%). So are they using the same number of chips ? If that is the case, it is hard to explain the servere drop in write speed.

 

wow that's pretty bad. Guess early adopters keep on getting burned pretty badly!

 

ok, found the official message in the mentioned article

Guide New update on the 25nm OCZ SSD drives

They do change to use 2x density die(64Gb vs 32Gb) which explains why the write speed drop by half. Same situation as in 80G vs 40V in Intel's case as the channels/blocks have been reduced by 1/2.

The only difference is that Intel made it crystal clear what you are getting, not so for OCZ. Not sure about other vendors of SF.

 

Yep, and thats why so many folks are up in arms over this. I was considering replacing my current drive with an OCZ. Going to wait and see how all of this plays out first.

 

I would forever ignore OCZ products unless they go bankrupt and start from new(with new management). A company with this management style cannot be trusted and a company you don't trust is not a company you want to deal with.

edit:
BTW, the 'explanation' of their official staffs about the reason is very fishy to me, telling you how sincere they are.

 

Wait...they only changed it since last week? I just got an OCZ Vertex 2 80 GB for my Latitude 13? Does that mean the 25nm SSD haven't been shipped yet?

 

80GB ? That sounds like a 25nm to me.

The raw size is usually 96GB(64+32) and they would use some of those as OP. Though taking a whopping 16 sounds a bit too much. puzzled.

 

Man now you are starting to scare me...though WEI is useless, it only rates my SSD @ 5.9 with latest IRST, when my old 30 GB Vertex on my i7 desktop is a 6.9.

When I go home from work, I'll run Crystal Disk Mark on it...

Is there a way to tell from the SSD the fab?

Edit: When the Intel G3 SSD come out, the Vertex 2 is leaving the Latitude 13.

 

I think they posted the detail model number of which is which. Though CDM with random data fill would tell you the answer as there is no way to dance around if the channels have been reduced.

 

Looking at the official announcement on the OCZ forums, the quick and dirty way to figure out what revision you have without opening the case is two things; firmware and formatted capacity. 1.28 = 25nm... 1.27 and below= 34nm. You will also notice an abnormally low formatted capacity. For the 120GB size I think you would be seeing 107-108GB now where you would have seen 111-112GB formatted before.

CDM might not help much unless you had to models side by side in similar condition. The only *sure* thing is formatted capacity and firmware version.

Oh and their trade-in program is a joke. Yea whoops we didn't tell you what drive you bought, but hey if you want to trade it in for the better version, just pay us extra.

EDIT:

It will be VERY interesting to see how this plays out for them with the FTC. These drives are being sold as higher capacity and faster speed than actually are.

 

Luckily, it only happens very recently so everyone can just use the 30 days return policy. They explanation of 34nm NAND spot price is sky rocketing is a very fishy way to explain why use higher density die when they LATER said the replacements would use same density die of 25nm. But knowing these guys(have a bit of conversation with them when I briefly had Vertex 2), that didn't surprise me.

 

Well 30-day window isn't available from everyone. Even big retailers like NewEgg will NOT offer any refund period on SSDs. 30-day replacement, no window for refund. You are stuck with it.

 

that is strange but I haven't dealt with newegg so I don't know. I returned my OCZ to tigerdirect without any problem(though not 30 days but more like 2days). Cost me the shipping charge but I am glad that it failed so early.

 

Yea depends on which place you bought it. Amazon you could probably return it melted in half from "fire damage" as long as you paid return shipping

 

Thanks Kevin, I will go home and look it up. I will post up my benchmarks later.

 

Even though this is a pretty new issue, a lot of customers have been complaining on their forum. It seems to me that OCZ forum admins are trying to make excuses. But, for whatever reasons, changing from 34nm to 25nm without any notice to customers was a bad thing to do.
Now a lot of customers would never go back to OCZ again I guess.

complaint about 25nm SSD drives

 

The smaller capacity drive also has a rated P/E cycle count of 3000, down from the previous 5000. So the SSD basically has 40% lesser longevity...and there is no way that an extra ~5GB spare scratch area is going to make up for that.

It is only the 2nd time they've silently changed the specs on their SSDs. The first time they dropped the P/E rating down to 5000, from 10000! They didn't pass on the cost savings then either.

OCZ is absolute trash, and I would never buy from them again. They don't support people who have problems with their OCZ RAM (with RMA times sometimes exceeding 60 days) and I've had three unstable SSDs from them that I was glad to have gotten rid of.

OCZ has a history of playing hardball. Storage Review was banned from OCZ's list of reviewers that got samples ( link), specifically because SR complained about drive stability and re-reviewed the SSD with OCZ-supplied updated firmware that had slightly lesser performance. Seriously?

The only time I had the pleasure of working with OCZ's customer support was when my two 60GB drives were unstable in RAID1. It was absolutely impossible that the SSD could be wrong! They pretty much refused to believe their SSD could possibly be imperfect. The last 'conversation' I had with them...they asked me about my RAM being a potential problem (also unfortunately OCZ hahaha) and after that I just got rid of the drives after firmware updates seemed to solve the stability issues.

 

those are not just OCZ forum admins, they are OCZ employee. A bunch of 'kids'(not age but they way they behave).

 

I figured this day would happen.

Typically after refreshes like 45-32 there is always a reduction in cost for the product, OR at the least the same cost but more space.

This is outright greed on behalf of OCZ. They reduce the amount of nand needed, increase the amount of back up nand needed, and keep the price the same. They have cut production costs, cut packaging costs, and kept the price the same while offering a technically inferior product.

 

I'm of the opinion that it doesn't matter how this plays out. It's all in the opening play.

I cannot recommend OCZ any longer.

 

I've stopped recommending them as soon as they stopped being the value brand.

 

Actually they do. It depends on what they are selling at the moment. To the point of almost bad mouth their own product(Vertex i.e. the one using Indilinx which of course is on the way out) just to show how good their Vertex 2 is. Telling you how they conduct their business.

 

I had been debating between Vertex 2 and X25M for quite a while. Eventually, I went for Intel because of its reliability. I am willing to give up a bit speed for that. Plus, I heard that OCZ's Vertex 2 SSDs recently changed from 34nm to 25nm which is slower. Also, some users noticed OCZ's performance decreases significantly after 3-4 months even with TRIM. Im not sure about that but I just went with Intel to keep my mind free.

You might wanna take a look at OCZ SSDs' reviews/feedbacks. These are for the old version 34nm. There are so many people saying their drives are either DOA or after couple of months. Totally unacceptable if that's true.

OCZ Vertex 2 120GB

Newegg.com - OCZ Vertex 2 OCZSSD2-2VTXE120G 2.5" 120GB SATA II MLC Internal Solid State Drive (SSD)
Amazon.com: Customer Reviews: OCZ Technology 120 GB Vertex 2 Series SATA II 2.5-Inch Solid State Drive (SSD) OCZSSD2-2VTXE120G

It cant be denied that OCZ's SSDs performed very well. On the other hand, its hard to prove they are reliable. I might not touch OCZ for a couple of years. And Im happy that I made a decent choice

 

This is ridiculous.

I can't justify what they've done to customers even if they all went on their knees and begged for forgiveness.

 

I bet this is gonna affect OCZ's Vertex 3 sales very badly. In addition, Vertex 3 might also be the last competitor of the next gen SSDs comes out. Hence, Intel clearly has advantage on the next gen SSD battle.

 

Well OCZ stopped selling RAM as of February, so they are going to primarily focus on SSD and power supplies. The profit margins on RAM were just too low, so they left the business. I think it was every 1 SSD sale, they had to sell 20 RAM to make money.

 

as if no one ever listened to me... i knew since years why i never want anything ocz.

 

Yeah very bad move not differentiating the slower NAND by packaging under Vertex 2 name. 25nm 64Gbit is up to 43% slower in writes than old 34nm used Result Mini-review: Direct comparison between 34nm and 25nm drives using 0-fill highly compressible data and 0/1-fill partially compressible data!

 

Ugh. While I'm happy with my Vertex 2 (which I got back in December, and has an old enough firmware (1.25) that I'm pretty sure it's 34 nm), I do agree that the way that they handled this has pretty much made for a PR disaster. What I'm more curious about is how much of all this is really related to the 25 nm NAND; after all, what if Intel is quietly doing the exact same thing with their G2s (meaning newly manufactured G2s have quietly replaced the 34 nm NAND with 25 nm), and it's just less noticeable since the performance of the Intels are low enough (comparatively) that the performance difference doesn't show up. This could mean that current Intel G2s have lost the reliability edge that the original G2s did. I don't expect this, especially after their Sandy Bridge chipset mistake, but it's an interesting speculation, none the less. The StorageReview article says that Intel plans to list "user capacity", but that's not necessarily helpful; if they reduce overprovisioning or just add a NAND chip to maintain "user capacity", that's not necessarily going to make up for any reliability or performance differences due to the change from 34 nm to 25 nm. So I guess the question really is what part of the problems are actually related directly to 25 nm NAND (as opposed to all the marketing problems that OCZ is having).

 

Jud, that's definitely a 34nm drive, the switch was last week and some retailers like Amazon even have the older 34nm drives still.

Intel is not doing this swap, I have confirmed that with them directly. No one else I have talked to is going to play this deceptive game. I've talked to 4-5 SSD companies thus far and expect more to report in today.

 

boys, i've been lucky i had changed my order of vertex2, but image those poor people without 30 days warranty return being screwed...

i bought ocz ram in the past and was good, but in theese day of aggressive capitalism at the edge of fraud, this is unacceptable..

boycott ocz, buy from other sellers, let them go bankrupt, pass theese info around web

 

An SSD controller is designed for the nand chips used: everything is inter-related and simply adding an additional nand chip is not possible because the number of channels is fixed for a specific controller.

To increase the number of channels, you need to double the capacity of the drive (for the same sized nand chips used). Of course, this is even over-simplifying it (controllers don't all scale so perfectly).

 

I just ran crystalmark on my OCZ vertex 2, and yup, I got pooched.

 

There is one born every second. Very few people made purchase decisions based on comments like here, all they need is some friendly reviewers(and there are lots of ways to get that)

 

Well, that's good to know. Part of the problem/worry is that the change was pretty much invisible, except for the lowered capacity on drives smaller than 240 GB (due to the loss in capacity from RAISE needing a bigger chunk of the higher density chips), and the performance differences. Or, I suppose, if you decided to crack open your drive and look at the actual NAND chips, but that kind of defeats the point (as well as the warranty!).

Well, it's definitely a bit more than that, I agree. For example, OCZ sells 60 GB Vertexes in 2 configurations, 1 with 8x 8 GB chips, and 1 with 16x 4 GB chips (source here, post #9). The thing is, until this came up, they were selling almost exclusively the 8x8 GB ones (to retailers; remember that OCZ doesn't sell direct to the end-user), because they were cheaper (and these are the ones that would have the performance differences, since I guess half the chips would mean half the channels... although I don't know if they use 2 different controllers for the 2 models). This doesn't prevent my reduction of overprovisioning argument, though.

It also doesn't seem to explain the performance drops for 120 GB drives, since when StorageReview did their review of the (old) 120 GB drive, it had 16 NAND chips, and by my previously cited link, so does the (new) 120 GB drive... which means the performance differences would seem to be related directly to the new 25 nm NAND.

Edit - a little further reading seems to imply that the difference between the 2 different 60 GB models isn't performance so much as just capacity; RAISE takes up less space on the lower density NAND, so there's more available space.

 

Yeah, but the old Vertex 2 itself had bad reviews - high failure rate. Because of that I did not purchase the OCZ.

 

It really is a shame. I've been mostly happy with my Vertex 1 drives. Still running strong for 2 years now. Didn't like the loss of destructive flash method which could be run from ACPI. Something important for most notebook owners since they can't switch to IDE mode in BIOS.

This debacle along with continued high prices will certainly temper SSD adoption rates. I hope Microsoft takes note and strives to maintain sub minute OS boot times for HDDs.

 

LOL I thought your nick comes from Vertex Pro 2.0

I agree. But that's how technology is. Expensive and buggy at the beginning, then becomes better and cheaper later. Just 2 years ago I could never imagine I can get 2TB HDD for $89

 

I gotta agree with you, my Vertex 1 30 GB has been going strong for nearly a year, though not the fastest no issues yet. I own 3 OCZ SSD and thankfully none of them have any issues yet but given what OCZ has now done, I don't think I'll be inclined to purchase another one.

 

Meh *shrug*

I don't think it's that this is too big of a deal. Mistakes in engineering & design happen. What matters to me is how OCZ addresses this, and what they do to make it right.

For example:

- Correct way: Intel's Cougar Point (Sandy Bridge P67 chipset) recall. Admit mistake early and up-front, stop shipments until problem is solved, issue a recall, and set aside a lot of money ($1B USD) to do right by the customers.

- Wrong way: Apple iPhone 4 Antennagate. Deny that there is a problem. Insist that it's the user's fault for holding the phone wrong. Show how several competitors' products also have the same problem. And also offer free bumper cases to customers, almost like you're impatiently and spitefully trying to get a bunch of crybabies to shut up. Yeah, not the right way to treat customers.

 

fun fact: we already know how ocz responded the last frew times. none of those where good responses.

so they cheat, then lie, then cheat again, then lie again.. how often do you want it?

 

Apple is quite possible the only company that can say 'turn around, bend ...'

 

Are these numbers worrisome? I have latest IRST installed for my chipset (GS45)

Newegg.com - OCZ Vertex 2 OCZSSD2-2VTX80G 2.5" 80GB SATA II MLC Internal Solid State Drive (SSD)

 

the sad thing is that they have benchmarks updated to new version, but the name, code product are the same of the old vertex2. So they are in the right, but everyone trusted in them buying with blind eyes, without checking their forum

 

thats why I have my intel 160G in my laptop, and put my vertex away

 

Thank god I got the original version 2 weeks back, but my next SSD purchase will be Intel G3 since they still seem to have better reliability.

 

OCZ has earned themselves a coveted spot on my 'banned' companies list. Right up with there with Xigmatek, NetGear, HP, Office Depot, and a few restaurants .

 

Man... some places never learn. The IHOP near work earned that spot for even my parents. Took like 45 minutes to get cold pancakes

 

Sounds like our local Braums... 20~ minutes to get a banana split. 30~ minutes for a burger (undercooked), etc. Horrible service, but funnily me and my sister came to a conclusions:
Out of ALL our restaurants, Braums is the most consistent. The only problem is that they are consistently horrible.