OCZ vertex 4 uses marvell controller

[bryanW1995]

You don't even know what you are talking about.

Marvell make the controller. OCZ buy the bulk of their NAND from Intel, who make it under their IMFT company.

So with that straightened out, yes, Marvell do make very good SSD controllers as the m4 and various others have proven. If OCZ released a Marvell based SSD with their own firmware like Crucial and Intel have, nobody would have said anything, just like they didnt then.

The fact is yet again they have given their customers the two fingered salute and let us to believe they had pushed the envelope in the SSD manufacturer world and had become an SSD controller and firmware supplier which would allow them to offer a more integrated and refined solution with better testing and validation. None of this is true.

This is the big issue for me. This is a complete PR flop for OCZ. I would call it a nightmare, but they've had too many of those already. Why didn't they just say that they were using a highly tuned (wink wink) marvell controller? Why the smoke and mirrors? As mentioned, marvell controllers have an awesome reputation, if anything this probably would have inspired more confidence from discerning buyers that OCZ was finally "getting it". Instead, all we "get" is that OCZ is filled with liars and crooks, who quite possibly could be lying just for the hell of it.

Right.

I actually would have felt better knowing it was a Marvell controller up front. How is that hard for OCZ to grasp?

Ride the M4/Crucial/Marvell good times wave.

The only thing that I can think of is that they were trying to differentiate themselves from a crowded marketplace. As usual, it backfired on them. OCZ is like the anti-Nathaniel Bedford Forrest: they could take theirs and lose to yours, or they could take yours and lose to theirs.

 

[bryanW1995]

Doesn't work? LOL What can I say to that? You want me to coddle you guys here? Kiss your boo-boo's even? Get over it already and save the thespian drama for the stage.

Fact is this. The drive (sometimes) works.

Why did you buy it in the first place?.. on controller design alone? What do you know about that controllers function or possibility that it would be so much better than the others? Err?.. wait.. do you even own one?

Here's how the bulk of people buy drives. Reviews and testing from other users. Now,.. does any of that hard data change after you find out that drive is a revised M4? Shouldn't matter one little bit if the drive is functioning as it was originally advertised.. unless you have something against Marvell, I guess.

This is just yet another bash-fest for those who like to watch boxing matches between the fanboys and the flamers. And to further that trend here?.. my guy has bigger knockout punches than your guy.

FTFY. It is OCZ, after all...

Based on the title of this thread, it can be construed in many ways. Saying that OCZ "scammed you".. especially when.. "you".. don't even own the damned thing, is really just preaching to the chior around this joint. Guess who the thread is going to attract?

I can only tell you that without a doubt here.. the Indilinx firmware is better than any other Marvell based drive I've tested so far(M4 and the 510). Obviously not as fast on the original Octane due to plain old nand.. but the on-the-fly garbage collection is top notch.

But if you don't beat the piss outta your drives with writes or use raid arrays without trim pass-through?.. then it's probably best to just be happy with what you have and don't split such small hairs.

So, until those who want to piss and moan about it.. can actually design their own SSD's and firmware code?.. you have to take what you can get in the current market.

If being told everything and knowing exactly what hardware and firmware code weaknesses you're getting is a requirement?.. good luck with that.

What is your definition of "better"? Faster? More responsive? Higher burst speed? More reliable? etc etc? From looking at Anand's reviews, it looks like the 830 is faster than the vertex 4 in real world scenarios, and from my ability to use the internet I've discovered that OCZ doesn't give a crap if their ssd's work or not. So, um, how is it better? It sure looks good at synthetics at least...

 

[bryanW1995]

Tbh the way the SSD sector has developed over the last 6 months I feel that Samsung is streets ahead of the competition.

I have high respect for Crucial for the way they have supported the m4, it's just a shame they never made a toolbox application for it and it is looking a bit long in the tooth now compared to SF and the 830 but it has always been very well priced.

Beyond that, I don't like how Intel treat their customers on their "community SSD forum" and clearly do not listen to customer issues with their SSDs. Also there has been too many reports of the 320 8MB bug continuing for me to feel as safe with it as I did with my old X25-M G2.

I've read around different places Intel 520 users have been reporting BSODs? We all know the score with non-Intel SF drives. Then OCZ have pulled another fast one with this Indilinx Infused masquerade.

FWIW, my wife has had no issues in 4 months of owning her 320 series. I did verify that it had the latest firmware before using it, however. The 330 looks interesting, but if it's a SF drive then I'll probably lean towards a couple of the marvell controller ssd's or maybe look for a deal on a couple of 830's.

 

[Coup27]

FWIW, my wife has had no issues in 4 months of owning her 320 series. I did verify that it had the latest firmware before using it, however. The 330 looks interesting, but if it's a SF drive then I'll probably lean towards a couple of the marvell controller ssd's or maybe look for a deal on a couple of 830's.

I have never had an issue with a 320 either. I have one in my laptop and 1 each in works 2 presentation rooms. Granted these are all very light duty machines. However if you look on Intels forum there are plenty of users who understand what they are doing reporting the 8MB bug after the firmware update and Intel have just ignored them. I don't like to see that. I also think the market for the "trusty reliable Intel 3Gbps drive" is shrinking fast. Before it was a safe heaven from dodgy Indilinx (pre-OCZ) or SF based drives. With reliable m4's and 830's available for the same or less money, I think the market for the X25-M and 320 is shrinking fast.

Regarding the 330, I don't understand that at all. If it is a SF based drive then it's basically a [somehow] throttled 520 for a little less money. That doesn't appeal to me one bit.

 

[Johnny Lucky]

There might be another way to look at this.

If one goes all the way back to March 2011 and reads the news articles about OCZ's acquisition of Indilinx it is perfectly clear that Indilinx was a research and development company with 45 employees and no fabricating facility. Indilinx still has no fabricating facility. In other words some other company manufactured the controllers for Indilinx.

There is no information anywhere on any official web site indicating either Indilinx or OCZ has a fabricaion facility for manufacturing ssd controllers.

If one visits the official Indilinx web site, the company clearly indicates it has a partnership with Marvell. One will find the following: "Everest SSD platform co-developed by Indilinx and Marvell".

Jump forward to the big CEBIT show where OCZ first showed off their new Vertex 4 ssd. Tom's Hardware noted the the Vertex 4 displayed Marvell characteristics.

A few days ago Anand reported and OCZ quickly confirmed the source of the Indilinx Everest and Everest 2 controllers is Marvell.

About the only thing one can safely say is the manufacturer of Indilinx controllers has been identified as Marvell.

This really is not much different from the practices of other hardware companies. For example, XFX sells power supplies. XFX does not design or manufacture power supplies. Their power supplies are manufactuered for them by Seasonic. AMD introduced a new line of memory. AMD does not design or manufacture memory. The AMD memory is manufactured for them by Patriot. There are hundreds of examples like that.

The only difference with the Vertex 4 is the point at which a component manufacturer is identified. For most components the actual manufacturer is identified during technical reviews. For some unexplained reason reviewers did not mention the manufacturer of the controller despite the fact that Indilinx which has no fabricating facility clearly indicated the controller was co-developed with Marvell.

 

[frostedflakes]

New account registered today. :hmm:

Anyway Anand's update suggests it's just an off-the-shelf (or slightly modified) Marvell controller. It's not just manufactured by Marvell, it was designed entirely by them and the Indilinx division at OCZ just wrote the firmware instead of designing the controller and writing the firmware. Which is fine, nothing wrong with that, but that wasn't what they implied earlier when claiming the design was completely in-house. Everybody knew they would have to send the design off to be fabbed, but people were expecting the design itself to be Indilinx's. Thinking back I don't know if OCZ ever explicitly claimed the controller was designed by them, but that's certainly the impression they seemed to give reviewers like Anand and they didn't bother to correct reviewers who passed those impressions onto their readers. Just one among the numerous deceptive and kind of shady things OCZ has done IMO.

edit: Marvell is fabless as well I think, so Marvell isn't even the manufacturer of the controller anyway.

 

[Coup27]

Just one among the numerous deceptive and kind of shady things OCZ has done IMO.

Something else which is shady is the above member registering their account today and making one post which is a very diplomatic and passive defence of both OCZ and Indilinx which ignores all of the points brought up by the users of this thread and puts the onus of blame on wrong assumptions made by the entire tech industry.

 

[Johnny Lucky]

Coup27 - You are absolutely corrrect. This is my very first post at AnandTech. I have been following AnandTech for quite a few years. This thread finally induced me to register and post a comment.

You are also correct that my post was diplomatic. I intended it to be so. However, it was not meant to be a passive defense of OCZ. It was not meant to put the onus of blame on wrong assumptions made by the tech industry. It was simply meant to provide some factual information and possibly a different way of looking at the situation.

FTR - I am only a new member at AnandTech. I am a Senior Member at Tom's Hardware. I have been for many many years. I use the same handle and avatar as here. If you go to the ssd section of the forum you will find an exceedingly large number of my comments in the threads.

I also maintain the ssd database listed in the in the sticky at the very top of the ssd section at Tom's Hardware. Here is the link to the database:

http://www.johnnylucky.org/data-storage/ssd-database.html

Finally, I am no fan of OCZ. I have never recommended OCZ products of any kind. Being a veteran poster my reason for not recommending OCZ is a bit different. Years ago PC Power and Cooling was known for high quality power supplies. I and other veterans did not hesitate to recommend PCP&P power supplies. When OCZ acquired PCP&C things changed. OCZ turned it into a hollow shell of what the company once was. Veterans might also remember issues with OCZ memory long before OCZ decided to get out of the memory business. It was just a matter of removing heat spreaders and identifying the actual memory components. Need I say more?

Peace!

 

[blckgrffn]

@ Johnny - impressive looking DB there

Although, somewhat ironically, it looks like you need to update the Octane & Petrol data

/has fond memories of the orignal OCZ powerstream, but that is all/

 

[Ao1]

There is nothing wrong with Marvell controllers or re-badging, but there is something very wrong with deceiving customers, reviewers and shareholders into thinking you are producing in house controllers when in fact you are re-badging someone else’s technology.

OCZ made some very strong claims about the benefits of using their own controllers to customers, reviewers and shareholders…..

Addition of Controller Technology is Expected to Substantially Increase FY 2012 Revenue and Margins
http://www.ocztechnology.com/aboutocz/press/2011/423

"Producing SSD controllers for OEM integration and use within our own products significantly enhances our ability to capitalize on worldwide demand for SSDs, and yields cost reductions which help make SSDs more accessible to potential customers," said Ryan Petersen, CEO of OCZ Technology Group.

"This combination puts OCZ in an advantaged position as one of the few SSD manufacturers with captive controller production. The acquisition will help to increase both revenue and gross margins as we vertically integrate controller technology, and it also provides critical support for current and future Enterprise and OEM opportunities."

Inclusive of Indilinx, OCZ expects its revenue for fiscal year ending February 28, 2012 (FY 2012) to be in the range of $300−$330 million, an increase of approximately 60%−75% from an estimated $189 million in its fiscal year ended February 28, 2011 (FY 2011).

As a result of the acquisition, OCZ expects its gross margin to increase by 2%−4% within 12 months of closing. Gross margins are expected to increase due to increased sales of controller products (which generally have gross margins in the 50%−60% range) as well as integration of Indilinx controllers into a greater range of OCZ products in the 12 months post acquisition. The Indilinx acquisition is expected to become accretive towards the end of FY 2012.

 

[N4g4rok]

If it addresses previous performance problems, what does it matter what they call it? What does it matter if the controller came from a different source? The issue is whether or not OCZ released a reliable drive. I wasn't aware of any of their problems when i got a Vertex 2 a year ago. Thankfully, its still working without a problem, but that could change next month.

If the new drive using Marvel controllers is more reliable and still outperforms Sandforce, then more power to them. The fact that they didn't really do their own homework immediately after the acquisition of Indilinx doesn't entirely effect the consumer. I feel like a small portion of the enthusiast market would be upset, primarily because they would rather see new ground covered as opposed to re-branding.

From OCZ's perspective, i'm not opposed to the idea of putting a Marvel based drive out to market that puts a small positive spin on the OCZ brand before they try something new.

 

[Ao1]

If it addresses previous performance problems, what does it matter what they call it?

But it doesn’t address previous problems. That is one of the core facts that have been misrepresented to end users, because the performance issues were allegedly due to the outsourced controllers that OCZ allegedly had no controller over. They claimed that having their own controller would give them 100% control.

Personally I don’t think you can just blame the controller. With SF the onus on vendors was to carry out validation testing, which was simply not done. Look at the time span between controllers being announced and products hitting the market and it is obvious that validation is being done in the field via paying customers. (This is true for both SF drives and the new Marvell based V4 product).

Have you noticed many problems being reported about the Intel 520? The 520 uses the very same SF controller as the Vertex 3.

Have you noticed many problems being reported about the Intel 510 or the Crucial M4? Sure, there have been a few niggles over the last year, but do they compare to what is reported for the Octane/ Petrol/ V4, which are using exactly the same (but allegedly) overclocked Marvell controller?

Check out the NewEgg reviews and OCZ forum and you will see a rash or reports about bad blocks, locks ups etc.

It seems OCZ can make a reliable controller unreliable, which makes you wonder exactly what they got for their $32M acquisition of Indilinx.

 

[Coup27]

Ao1's post says it all. Another new member who has joined this thread lol.

It clearly states there in OCZ's own quote that they've peddled the benefit of in-house controller facility.

To the people saying "does it matter", unfortunately I believe you have missed the point of this thread. The point is not "they released a reliable drive, whats your problem" it is the smoke and mirrors that has gone on to conseal it wasn't theirs all along. Had they said "We're working on our own controller but our next gen drives are based on a Marvell design" nobody would have said anything.

 

[Ao1]

Had they said "We're working on our own controller but our next gen drives are based on a Marvell design" nobody would have said anything.

Exactly.

And to those saying “they released a reliable drive, what’s your problem” I would say are you sure it is reliable? It was rushed out the door and a lot of compromises were made in the process. Plus at the 128GB capacity range (which is the range that most people buy) performance sucks.

 

[Obsoleet]

I have never had an issue with a 320 either. I have one in my laptop and 1 each in works 2 presentation rooms. Granted these are all very light duty machines. However if you look on Intels forum there are plenty of users who understand what they are doing reporting the 8MB bug after the firmware update and Intel have just ignored them. I don't like to see that. I also think the market for the "trusty reliable Intel 3Gbps drive" is shrinking fast. Before it was a safe heaven from dodgy Indilinx (pre-OCZ) or SF based drives. With reliable m4's and 830's available for the same or less money, I think the market for the X25-M and 320 is shrinking fast.

Regarding the 330, I don't understand that at all. If it is a SF based drive then it's basically a [somehow] throttled 520 for a little less money. That doesn't appeal to me one bit.

I concur with this, if the 330 is a throttled SF 520 then I don't see the point to its existence. A BSOD prone, slower 520? No Thanks.

The reports of BSODs on the 520s are too widespread for me to buy one or recommend one. The market for reliable 3gb Intel drives like mine is, or should be, gone. Mine work great and I can't justify replacing them until I move to a 6gb system, but this doesn't change market realities. I'm not one to simply promote what I bought (in this case I purchased 3 to 4 years ago).

I see no reason to disregard most choices and just pickup a cheap M4 or just go all out and get the 830. If Crucial would develop a toolbox for the M4, they'd be the new "G2".

 

[N4g4rok]

Have you noticed many problems being reported about the Intel 510 or the Crucial M4? Sure, there have been a few niggles over the last year, but do they compare to what is reported for the Octane/ Petrol/ V4, which are using exactly the same (but allegedly) overclocked Marvell controller?

I thought that each manufacturer makes changes to the controller, which causes the difference in SSD reliability/performance. By that i meant SF controller in the 520 would not be exactly the same as the on in the V3.

To the people saying "does it matter", unfortunately I believe you have missed the point of this thread.

I think you're right on that. Ended up in a strange direction with that post. Whoops.

 

[Ao1]

The reports of BSODs on the 520s are too widespread for me to buy one or recommend one.

Not doubting you as I’m sure someone somewhere will have reported a BSOD, but widespread? Can you give me a few links to demonstrate the diversity of the problem?

 

[Coup27]

tbh mate a quick google of Intel 520 BSOD brings up loads of links all over the place

 

[Ao1]

Hhmm, a quick google/ bing search didn't seem to reveal many reports at all, although I'm sure they do exist.

AFAIK the SF controller is exactly the same in the V3 and 520. The f/w is however different. Intel undertook a lot of validation and bugs in the firmware were fixed as a result. Consequently the 520 is quiet reliable, whilst on a comparable basis the V3 is not.