The source of OCZ flash

[autosax]

Hi guys,

I heard some problem about OCZ Flash source, below is the link, but it was from china website so please use translation.

Link

they are talking about OCZ flash is not good.
:thumbsdown:


Locking this as thread has run its course.

n7
Memory/Storage Mod

 

[autosax]

Guys, check this link, you will say bad words and angry like me.
:thumbsdown::thumbsdown:

from OWC BLOG

 

[taltamir]

Generally "check this out" and a link are taken as advertisement / spam. In the future you might want to flesh out your argument a bit more.
However your two sources are very different and unrelated sources so I do not think that was your intent.

Anyways, I have been reading through them and it is interesting, and concerning info.

The truth is, we did not set out to see whether OCZ had cleared up the capacity discrepancy. Rather, the original inquiry was to determine if these retail sourced SSDs shared a troubling concern we uncovered with the third SSD, a 240GB Vertex 2 model, obtained directly from OCZ.

When we took the cover off of this third, direct from OCZ SSD, we found a ‘S’ stamped over Micron logo on all the flash devices (see the image to the left). This indicates the device is “off spec” product because it failed some parameter of Micron’s full performance and/or quality specification testing. “Off spec” memory is typically used in low-level applications such as toys, offering considerable cost savings over Tier 1 level to an SSD manufacturer.

Tier 1/Grade A will have the logo of the manufacturer…such as Micron, Intel, Hynix, Samsung clearly printed/indentified on the chip.
All flash memory is created from one large silicon wafer. The best chips, Tier 1, come from the middle of the wafer while those cut closer to the edges of this wafer tend to be of lower quality.

OEM level chips don’t have that logo….but are considered very suitable for all full spec applications.

non-spec memory is rejected or “fall out” memory from the full spec….it failed some testing or quality control and thus should not be utilized in full spec applications.

If this is legit then it is ridiculously bad. However, I am not ruling out FUD yet considering the sources.

 

[autosax]

Hi taltamir,

Sorry for my bad English.
I am not professional at SSD.
I just know Flash will effect SSD life and performance.
We can see OCZ use original flash for review.
I think it is a problem.

 

[Emulex]

That's like saying corsair/G.skill doesn't use some of the bargain basement stuff (1.65v ram) - duh.

There is no way that they can get Micron tier-1 - Intel/Micron have that locked down. but if they want some scraps - of course.

I thought you guys knew how binning worked?

 

[nanaki333]

i'm certainly no reliable source, but i've had nothing but bad luck with ocz branded drives in the flavor of vertex 1 & vertex 2 drives.

 

[spooky69]

http://www.ocztechnologyforum.com/f...nm-Spectek-Nand...see-here-for-specs-and-info

 

[BTA]

I'm sure it's the cheapest shit they could get their hands on and slap their name on it.

 

[=Wendy=]

SpecTek parent company is Micron.
From what I understand.
Micron keep the NAND chips from the center of the wafer, and the rest goto SpecTek for binning.

 

[Denithor]

I'm sure it's the cheapest shit they could get their hands on and slap their name on it.

+1

OCZ would accept memory that meets their requirements, whether or not it meets the Tier 1 requirements of Micron/whoever. They then build the SSD and ensure it will meet whatever specification they want to print on the box.

If you test and it doesn't meet this spec - just RMA and try another.

 

[spooky69]

SpecTek parent company is Micron.
From what I understand.
Micron keep the NAND chips from the center of the wafer, and the rest goto SpecTek for binning.

That is not the explanation that I have read.

I don't see anything wrong with sourcing as cheaply as possible as long as the quality, reliability and performance is in line with the stated specification. I believe that OCZ offer a 3-year warranty and that they have been quite clear in stating the quality that the sourced chips have, so I do not see this as something that would influence my buying decision.

If the explanation given by OCZ is correct and if it is the case that OWC have ordered an OEM part and then misled people by the wording of their blog I think that this is likely to make me not want to buy from OWC.

 

[=Wendy=]

I don't see anything wrong with sourcing as cheaply as possible as long as the quality, reliability and performance is in line with the stated specification. I believe that OCZ offer a 3-year warranty and that they have been quite clear in stating the quality that the sourced chips have, so I do not see this as something that would influence my buying decision.

I don't see anything wrong with it either.
The SpecTek NAND that is being used is AL spec, according to OCZ, which is spec or above spec, so could in actual fact be better quality than Micron OEM NAND.

It would appear that whoever made the OWC blog entry, doesn't understand how binning works.

 

[taltamir]

the issue in question isn't that OCZ isn't shipping tier 1 ram.
The links say that some of their drives are OEM (tier2?) unbranded units... they pass QC but are not the best of the best parts. This is what is supposedly used by most manufacturers.
But then some of the units supposedly actually ship with NAND that has an S stamped on it which supposedly means "out of spec" aka "failed validation". This is the issue in question.

The expected thing would be for them to use OEM, its not tier 1 but it is in spec. Now the question is, do they take those failed spec ram, perform their own high quality validation (I somehow doubt that) and only use the ones that are still realiable (maybe their out of spec value is not critical)... or do they just stick them in and hope for the best.
I am really not sure what to make of this, is this really as bad as it sounds?

 

[nanaki333]

Now the question is, do they take those failed spec ram, perform their own high quality validation (I somehow doubt that) and only use the ones that are still realiable (maybe their out of spec value is not critical)... or do they just stick them in and hope for the best.
I am really not sure what to make of this, is this really as bad as it sounds?

right now, i wouldn't put it past ocz. they tried to pawn off drives with half the performance for the same price when they switched to 25nm. then, originally, had the nerve to charge people to swap them for 34nm drives. it would also explain why i see more bad eggs on newegg for ocz based drives than any other brand of the same type of drive.

 

[mosco]

Can anyone explain how the process works? The AL seems to be stamped in the same fashion that the micron logo is, not the way the S is. That would seem to imply that it wasn't good enough for micron although stamped AL and then it was sent over the spectek.

 

[MagickMan]

Doesn't surprise me, OCZ is a pathetic excuse for a tech company.

 

[taltamir]

Can someone please confirm/refute that an S stamp from IMFT means out of spec as the articles say?

 

[groberts101]

I have purchased more than 40 OCZ SSD's and all were as advertised. I still own 15 of thier SSD's and have not one single issue with them. All the crap we hear about this is purely a systematic "chumming of shark filled waters" aimed at turning some potential customers away until the "spin docters" newer models are released.

As for the "more bad eggs on newegg for ocz based drives than any other brand of the same type of drive"? Well.. considering the sheer volume of drives sold compared to some of the competitors, it would make logical sense. This I do know for fact. Take a look at other mfgrs support forums and you see mirror image issues in regards to the Sandforce controllers compatibility with laptops and some mobos bios implementations. Has nothing to do with a particular mfgr since they are subject to Sandforces limitation in this regard. Fact is that the drives rarely "die" and are simply panic locked which requires RMA since Sandforce would never let a destructive firmware into the hands of consumers. Too much NDA and shrouding by Sandforce has caused headaches for ALL who use them, INCLUDING OWC.

If you guys smearing and complaining really understood what your talking about here you'd realize that no one actually mfgrs these things in house. They surce the parts and have them outsourced/assembled to specs. This includes firmware to a certian degree, but they still have strict limitations due to Sandforce's slow testing and bug fixing timelines. IOW, IT'S ACROSS THE BOARD and boils down to the controller.

You can't simply say the drive is crappy because your Dell laptop locked the drive when it boils down to controller/firmware based issues since that falls squarly on the bios implementation of the system mfgr and Sandforce themselves.
Although, I too am tired of many companies strategizing to get the money first and fixing the issues after the fact some of it just boils down to trial and error. But hey.. that's the good old capitalistic tendency to please the stockholders taking affect there and is also across the board and very much non-mfgr specific as well.

Fact is that ALL SSD's do NOT work on ALL hardware. Sandforce controllers just seem to have more issues than most others in that regard. Talking about nand quality just shrouds the real issue and puts more chum in the water to call back the sharks that have been slowly moving away due to OCZ stepping up to take the heat for thier mistakes with the 25nm nand debacle. VERY few have been charged for the RMA of these drives and the few that were have been promptly refunded for the cost out of pocket. Many drives have also been shipped to remote locations in distant countries at full expense. Surely a company that has actually taken a loss on a particular products profit margin, can't be all bad? I'd like to see OWC try to step up and take that hit. They'd probably just file for bankruptcy and call it a day. LOL

PS. the OP of this thread has been on some type of campaign here with multiple postings across a few different forum so far and seems to go far beyond simple curiosity of this subject. Usually, if it smells like a rat.. and walks like a rat....

 

[taltamir]

I have purchased more than 40 OCZ SSD's and all were as advertised. I still own 15 of thier SSD's and have not one single issue with them.

Completely irrelevant to the issues at hand.

All the crap we hear about this is purely a systematic "chumming of shark filled waters" aimed at turning some potential customers away until the "spin docters" newer models are released.

No, its legitimate grievances.

As for the "more bad eggs on newegg for ocz based drives than any other brand of the same type of drive"? Well.. considering the sheer volume of drives sold compared to some of the competitors, it would make logical sense.

A logical argument. If over 50% of people only vote when there is a problem, then the higher volume you ship the more negative your score becomes.

You can't simply say the drive is crappy because your Dell laptop locked the drive when it boils down to controller/firmware based issues since that falls squarly on the bios implementation of the system mfgr and Sandforce themselves.
Although, I too am tired of many companies strategizing to get the money first and fixing the issues after the fact some of it just boils down to trial and error. But hey.. that's the good old capitalistic tendency to please the stockholders taking affect there and is also across the board and very much non-mfgr specific as well.

Fact is that ALL SSD's do NOT work on ALL hardware. Sandforce controllers just seem to have more issues than most others in that regard. Talking about nand quality just shrouds the real issue and puts more chum in the water to call back the sharks that have been slowly moving away due to OCZ stepping up to take the heat for thier mistakes with the 25nm nand debacle. VERY few have been charged for the RMA of these drives and the few that were have been promptly refunded for the cost out of pocket. Many drives have also been shipped to remote locations in distant countries at full expense. Surely a company that has actually taken a loss on a particular products profit margin, can't be all bad? I'd like to see OWC try to step up and take that hit. They'd probably just file for bankruptcy and call it a day. LOL

completely unrelated to anything discussed in this thread.

PS. the OP of this thread has been on some type of campaign here with multiple postings across a few different forum so far and seems to go far beyond simple curiosity of this subject. Usually, if it smells like a rat.. and walks like a rat....

please avoid attacking the OP and other posters, if you have something to say that actually relates to the topic discussed in the thread itself then please share it. If you believe someone to be trolling then report the post.

 

[belladog]

I agree OCZ should be hauled over the coals for this. Its dishonest, and its only a few weeks after they got busted on the 25nm nand issue so they have form.

If you're buying a $20 memory kit you would expect cheap oem parts but when a $120GB SSD costs a few hundred dollars we expect tier1 components. The bottom line is they are using the marketing and branding of a A grade components, then selling cheaper less quality nand to the masses for the same price.

They continue to sell 25nm vertex2 drives under the same branding when other manufacturers like Intel and corsair are doing the right thing and rebranding the drives.

They have always been secretive about which nand is used in which drives and now we know why.

@Groberts, since you get free drives from OCZ i'm going to take your comments with a grain of salt.

 

[autosax]

I just curious, why only OCZ had those strange problem (25nm issue, flash??maybe).
Does OCZ think users are blind.
:hmm:

 

[Emulex]

all i know is the crapton of intel drives have been giving me less trouble than the variety of sandforce - so i hope indilinx can bring some reliability back to the game.

everyone knows Sandforce is plagued with bugs - to no end. Nobody can fix all of them. They do not exist in other chipsets. that is enough to fail.

Throw "nand du jour" sales on top of that with freaky sizes and special editions; it's pretty obvious.

This is the seagate 7200.11 meltdown of OCZ

 

[groberts101]

Completely irrelevant to the issues at hand.

No, its legitimate grievances.


A logical argument. If over 50% of people only vote when there is a problem, then the higher volume you ship the more negative your score becomes.


completely unrelated to anything discussed in this thread.


please avoid attacking the OP and other posters, if you have something to say that actually relates to the topic discussed in the thread itself then please share it. If you believe someone to be trolling then report the post.

If the context of my reply is related to those making slanderous remarks about the 25nm debacle are any indication of the "looseness" of this specific topic, then they are along the same lines. We are talking about nand specs here right?

I simply make observations of what I see and test and you guys seem to be doing the same so my remarks are no more irrelevant than anyone else here in this matter.

As for being biased for receiving free drives? I beta-test the new V3 and was actually very surprised I was invited to do so. I've gotten more than a few threads shut down over on OCZ forums due to my "UN-biased consumer type remarks". I hold no allegiance to them and simply try to help others with issue on their systems as a result of using OCZ drives and learning many hard lessons myself.

Let's be perfectly clear here. When there are millions of dollars at stake.. ALL mfgrs will take advantage of some situations. Not saying in the least, that anything in regards to the 25nm fiasco was justified, but that's big business in general and many times they continue on fully aware that they will come out ahead even after class actions are tallied up. All the other mfgrs names thrown around here have done things that are less than ethical or widely accepted by everyone too. Heck... Windows operating systems alone.. is probably responsible for more data/$$$ loss than all others combined but yet we still buy the next version hoping for a better experience.

I buy based on performance and everything else becomes secondary. Sandforce controllers give you that in spades and I could really care less if the nand is grade 12 or whatever many are trying to play it off as, so long as I'm not constantly having to RMA my drives as a result or losing performance in the process. Most of the issues you guys complain of or make reference to are simply compatibility issue as described earlier and have absolutely NOTHING to do with nand quality. Panic locked drives are a controller mfgr issue not an SSD vendors.

Seems to be a lot of bashers around this joint and gets pretty tough to sort the haters from the one's trying to help others with opinion and fact. I don't insinuate that I know everything there is to know about SSD or any specific mfgr and simply feel that others should stick with what they know and try to do the same.

As for "calling out" trolls around here?... the attention I attract with my posts tends to do that on its own without the need to run around tattle-telling. And if I get in trouble for it.. well.. that wouldn't be the first time. :sneaky:

 

[kingtaro]

groberts101

All Sandforce drive performs identically the same.Why users choose OCZ?? I believe it's pretty much because of the competitive pricing and the brand.

spectek is known as Micron's subsidary to sell downgrade chips, no matter dram chips or flash chips. Tier1 one wafers go to crucial or IMFT, crappy ones go to spectek and sold with very low price. Most of these chips are sold to emerging markets, such as china etc..

OCZ markets itself as high quality and high performance SSD vendor, but it uses downgrade flash chips inside. The relibility is the biggest concern as many users may suffer data lose or the drive just die easily.

Every company has their own definition of " high quality" That's why OWC mention this sendtence at the end of the video "It may be enough for OCZ, but not for us"

 

[taltamir]

why does everyone keep on saying "tier 1 vs OEM"
The articles linked make clear that there are 3 seperate tiers.
Tier 1 - top NAND
OEM - Ok nand
S - NAND that failed validation and is suitable for nothing but toys.

The issue isn't that OCZ uses OEM class NAND, that is a given. The issue is that they use S class NAND in SOME of the drives (others are OEM).
If you have verification that the articles are full of it then say so. but "OCZ using OEM is normal" type of responses that people make in this thread again and again are completely unrelated to the actual issue being discussed.