* Do not buy: OCz Vertex 2 and Agility 2 - Beware!

[deeppow]

It's not the NAND chips themselves affecting performance. OCZ used 8x8GB NAND chips in the 24nm version instead of 16x4GB in the 32nm version. This means the SF controller only has 8x potential memory lanes instead of its max 16. However, OCZ is allowing people with affected drives to trade up to a 24nm 16x4Gb version for free.

That probably explains the reason for the slower writes. Why do we loose storage space? That isn't apparent to me.

Seems like any semi-intelligent person could have guessed the resulting slower writes (maybe some potential affect on reads) yet they didn't test (change) the specs.

The specs, there in lies their problem.

 

[deeppow]

Thanks Dooms and deeppow I was trying to piece together what little i knew about ssd and several treads I had read. Is there a link between the change in NAND chips and higher failure rates that some people are commenting about or is that "unrelated"?

I don't know. Longer term testing upon which warranties are based is always a problem.

 

[Dooms101]

That probably explains the reason for the slower writes. Why do we loose storage space? That isn't apparent to me.

Seems like any semi-intelligent person could have guessed the resulting slower writes (maybe some potential affect on reads) yet they didn't test (change) the specs.

The specs, there in lies their problem.

I've read the storage thing explained several times by several and it still doesn't make any sense to me. Bottom line is I am officially never buying OCZ again. 3 reasons I've had to not buy from them:
1.) This false advertising crap.
2.) 2 Bad RAM kits from them.
3.) Jipped me on an RMA...

 

[bing]

After reading all these development and this particular post HERE, its quite clear OCZ was sitting at the complacency chair for too long.

Its just the greed, someone must be thinking its too easy to reap easy money by "quitely" slipping in the 25nm flash at the current Vertex 2 product line that is selling like hot cakes without people noticing it, looks like the decision maker was underestimating the crowd.

Damage control at this stage is quite difficult and the "price and cost" to fix this are definitely really high, I guess its way too high compared to the gain they expected to get from this move.

Really bad decision OCZ, really bad.

 

[I.M.O.G.]

Latest update added here:
http://www.overclockers.com/oczs-25nm-vertex-2-performance-issues/

Anyone with one of these that is unhappy can get it replaced. OCZ pays shipping both ways, and no cost for pricing differences.

 

[deeppow]

Latest update added here:
http://www.overclockers.com/oczs-25nm-vertex-2-performance-issues/

Anyone with one of these that is unhappy can get it replaced. OCZ pays shipping both ways, and no cost for pricing differences.

And my time and system down time? The costumer eats that of course. As far as I'm concerned that FAR exceeds any value of the replacement.

Invest the $s you made from this wisely OCZ, you'll never get another one of mine.

 

[I.M.O.G.]

You are screwed on your down time.

I don't need to make excuses for OCZ, but in my experience talking with them they are a pretty cool company which is why I just look at this as a screw up and it happens. Other companies have made similar mistakes and gotten burnt in the past also. Poor judgement and foresight.

This situation is not a good one - they screwed up. I am glad however that they are doing what they can to address it at this point. I would think they have taken a lesson, as this definitely has grabbed their attention, and hopefully mistakes like this aren't repeated in the future.

 

[bing]

Just curious, since I'm not US citizen, is there a possibility that this issue will turned into a class action lawsuit by a group of unhappy owners, even OCZ provided no cost to replace them ?

Should this post way off topic, feel free to delete it.

 

[Jmtyra]

Anything is possible, as people are fairly sue happy here in the US. While OT, I'm doubtful your residency would affect your ability to participate in such a legal engagement should one occur. However, I'm not a legal expert, and could be completely wrong here.

 

[updawg]

Does this effect me? Here is the part number off my agility:
oczssd2-2agte0g

I only saw the vertex being mentioned. Thanks.

 

[5er driver]

Well, I'm glad I purchased the G.Skill Phoenix instead of the vertex II this time around I guess! Thanks G.skill for being $5 less....LOL

Like I said earlier I was about to get a Vertex 2. I'm now looking at the G.Skill Phoenix and see that its specs are almost identical to the Vertex 2. Are we sure that the Phoenix is still being shipped and offered with the 3xnm nand? If I can find out for sure, I'll get one immediately. Or, can anyone suggest a comparable SSD with 3xnm nand in the 60 to 80 GB range with reads/writes around 285/275 MB/s?

 

[bing]

Like I said earlier I was about to get a Vertex 2. I'm now looking at the G.Skill Phoenix and see that its specs are almost identical to the Vertex 2. Are we sure that the Phoenix is still being shipped and offered with the 3xnm nand? If I can find out for sure, I'll get one immediately. Or, can anyone suggest a comparable SSD with 3xnm nand in the 60 to 80 GB range with reads/writes around 285/275 MB/s?

GSkill is still using 34nm for their ssd, fresh info from the GSkill staff -> HERE

 

[Nycatt]

I was looking to get the Vertex 2 200gb version. Should I not? What is recommended?

 

[Dooms101]

Maybe I should point this out again... It is NOT the NAND size (24nm vs 34nm) that makes the difference here. It is the amount of IC's used. It is possible that other companies made 8 IC SSD's and these would be experiencing the same decrease in performance. SSD's with SandForce SF1200 controllers have a max 16 memory lanes and only SSD's with 16 IC's will have the max performance.

 

[Just a nickname]

Frontpage post:
http://www.overclockers.com/oczs-25nm-vertex-2-performance-issues/

My god! they keep their victim position:
"Regarding the performance, these drives are able to meet our published specifications with the same benchmarks we have always used, which is ATTO for maximum rated sequential speeds and IOMeter for random IOPS. We have also recently added a dual spec to show performance in ASSD on our product pages, we realize every benchmark is different and handles data differently."

 

[ratbuddy]

Uhh, you don't have to yell, and I already made that clear in the front page thread

 

[deeppow]

I don't need to make excuses for OCZ, but in my experience talking with them they are a pretty cool company which is why I just look at this as a screw up and it happens. Other companies have made similar mistakes and gotten burnt in the past also. Poor judgement and foresight.

This situation is not a good one - they screwed up. I am glad however that they are doing what they can to address it at this point. I would think they have taken a lesson, as this definitely has grabbed their attention, and hopefully mistakes like this aren't repeated in the future.

Thanks for your comments. A few related (and unrelated) thoughts:
First, most organization have some very good people (cool folks) in them. And that will certainly be true of OCZ.

Second, it is a little beyond "poor judgment" nearer to dishonest as far as the product they delivered versus their advertised specs. I guess the FTC could get involved if someone bothered to fill a complaint but I doubt it.

Third, if they had stepped up to the plate from day one to address the problem as Intel and the mobo makers for the Sandybridge quickly did, while there might have been some grumbling most would have worked through the issues. However that hasn't been the case, instead OCZ has retreated to their current position -- the one they should have taken to start with.

 

[Dooms101]

Uhh, you don't have to yell, and I already made that clear in the front page thread

I know, but apparently some people can't read and are still confused.

 

[jmh547]

Does this effect me? Here is the part number off my agility:
oczssd2-2agte0g

I only saw the vertex being mentioned. Thanks.

I read somewhere (can't find it again) that anything with an 'E' in the model it has the 25nm chips.

 

[Dooms101]

I read somewhere (can't find it again) that anything with an 'E' in the model it has the 25nm chips.

Not true, most with the "E" are the 8 IC versions, however some from before January are the better 16 IC ones.