OCZ Technology Vertex 3 SSD 120GB and 240GB Review @ [H]

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,533
OCZ Technology Vertex 3 SSD 120GB and 240GB Review - Today we take a look at the OCZ Technology Vertex 3 in both 120GB and 240GB capacities. The Vertex 3 was the first SF-2000 based client SSD to hit the market in 2011 and such will be the product in which all others will be judged against. The Vertex 3 is capable of transferring data at 550+MB/s.
 
Page 3 large images are not attached in the article. Working on getting that fixed. Thanks for your patience.
 
I had a Vertex 3 240gb MAXIOPS on pre-order with amazon for about two months, it has never been in stock on amazon and I have never seen it on newegg or any other major retailers.

Personally, I gave up last week when I was able to get a M4 256GB for $380, but if anyone knows where to find the 240GB MAXIOPS, let us know.
 
Note that the current crop of sandforce ssd's (not just OCZ) seem to be causing random BSOD's for some. OCZ issued a new firmware that supposedly is a workaround (not a fix yet), but verdict is still out on its effectiveness.
 
Nice review, thanks for putting it up. It's cool to see you guys getting back into SSD testing (actually IDK if you ever did).

The mistake they made with the enclosure really is inexcusable. They didn't measure the prototype enclosure to make sure it conforms to ATX specification? Why the hell not?

Even if a 650MB/s drive comes out six months from now, you’re really not going to care since we are approaching a point where the speeds are not showing up in real world use. If you are trying to open something that is 300 MB in size it doesn’t matter if you read 500 or 1000MB/s, the file will open in less than a second.

This part kind of puts me off regarding the newest SSDs. Aren't the SF-1222 drives of last year in this realm too? I just think that when you can get 95% of the maximum possible performance for 50% of the price, you're an idiot to pay twice as much. Granted, SF-1222 isn't 95% as fast as these nice Vertex drives, but my point still stands: if you're an average user or enthusiast, I hardly see the point of this.

Thanks again for the review.
 
Thanks for another great article Kyle! I just dropped a 25nm 60gb Vertex 2 in my rig (picked it up on sale for $80) and have been EXTREMELY happy with the performance compared to my older 640gb WD Caviar Black (now my storage drive). I can't quite imagine the idea of loading into windows before the splash animation finishes, but its cool to see the performance you can get now.
 
There is an issue sometimes with GBT boards and installing Windows from a thumb drive and that sounds a lot like your issue mashie.
 
Note that the current crop of sandforce ssd's (not just OCZ) seem to be causing random BSOD's for some. OCZ issued a new firmware that supposedly is a workaround (not a fix yet), but verdict is still out on its effectiveness.

A quick note on the BSOD thing. There is an issue but I haven't seen it so it isn't effecting everyone. This article was written more than a week ago and in the time the article matured the issue gained some traction in forums and with OCZ / SandForce.

I have 8 SF-2281 drives in house right now and 4 of them deployed in daily use systems. I haven't been able to replicate the issue at all. I actually haven't had a single issue other than the two mentioned in the article, casing form factor and Agility 3 performance when data is present on the drive.

At this point I'm in the same boat as you guys, no real answers and waiting for someone to fix something that I personally haven't experienced.
 
Last edited:
I got the Vertex 3 240GB drive a few days ago. Unfortunately it is stuck on the Marvell controller on my GA-Z68X-UD7-B3 as the Intel (6G) controller just refuses to play along. I will reach about 60% through the expanding of the Win 7 files during install before it just stops doing anything at all.

I had the same issue with a Corsair Force Series 3 120GB drive but my old Intel X-25M is working just fine on the Intel controller. Both the Vertex and the Corsair are working just fine on the Marvell controller... :confused:

The only thing in common between the drives that have issues are the Sandforce controller and that they are 6G.

Any ideas?

The unifying thread for people with BSOD issues with the SF-2281 drives is that most, if not all, are using the Intel 6GBps controller (P67/H67/Z68). I was one of them - I had an Agility 3 for about three weeks shortly after it first came out, and nothing I did made the BSOD problem go away. The tech support guy, who I was being nice to, basically told me something was wrong with my system, it wasn't the drive, and I should just return it to the retailer if I wasn't capable of figuring out what the problem was. When I politely informed him that I couldn't do that because the retailer didn't accept refunds, his words were, and I quote, "Not our problem. Ask them if they can make an exception." *click*

And all I did was tell him what the drive was doing and ask if I could RMA my drive or if they had any known fixes.

After OCZ basically told me to go sit on a shaft, I did manage to convince Newegg to let me return it for credit in spite of their no-refund policy on SSDs. I got an Intel 510 and *gasp* it worked without any dicking around.

Meanwhile, the complaints over on the OCZ forums all showed the same symptoms. OCZ went weeks saying it was not a problem with the drives, going so far as to differentiate their drives from Corsair's (who issued a recall for exactly the same problem). The growing multitudes that hung on went through multiple secure erase and new firmware cycles only to discover that not only did the BSOD/detection problems not really get sorted, now some drives are showing 50% (or more) performance degradation with the newest firmware. OCZ's reaction? It's your overclock. And your cables. And the fact that you're using your computer. And because you smell funny. Because it's certainly nothing to do with the drive, unless enough people complain.

To be fair, Corsair is having similar problems with the Force 3, even after the recall/rerelease and delayed the Force GT for the same reason. People are pissed off there too. But at least they're admitting that there's a problem and are sending out a mea culpa instead of telling everyone on the board to piss off.

I got burned once by OCZ around 2002-ish when they sent me and practically all of their customers RAM that didn't work at anything close to their specifications, and then magically disappeared when anyone tried to RMA anything. Remember that? IIRC Kyle and Anand both nearly got into legal fisticuffs with Ryan Petersen over that shit.

Now it's 2011, and I nearly got burned by them again. (A friend of mine literally got burned by an OCZ power supply that burst into flames last summer ... and which OCZ refused to RMA because they had forgotten to put a warranty sticker on it. But that's a different story.) To me, it's still the same shady-ass company as it was 10 years ago, just bigger and with more industry clout. They seem to only take care of their customers when the public eye is on them.

Anyway, I'm way off topic. Caveat emptor to anyone buying any SF-2281-based drives for Sandy Bridge platforms right now, OCZ or not. It's a roll of the dice as to whether your system will puke it up or run fine. Good news is that Sandforce is apparently aware of the problem and it sounds like more firmware iterations are forthcoming.

</rant>
 
I should clarify that not all people running these drives on Cougar Point are running into the problem. It's just that those that are almost uniformly are a) running Z68/H67/P67 and b) running off of the Intel 6Gbps port.

The temporary workaround for the problem is to run them off the 3Gbps port until they've got a fix (that's the official word from Corsair; OCZ is silent on the matter last I checked). Also interestingly, the vast majority of issues, anecdotally anyway, are with the 120GB drives, but that could be because those are the volume sellers rather than any positive correlation.
 
The Corsair recall is NOT the same thing that you guys are running into. I really can't go into details about it but look at my title above and understand that I hear unofficial things from official people and sometimes I'm given answers that aren't meant for public consumption.

I will just say that the Corsair issue is not the same as the other issues. That said, look at how quickly Corsair reacted.I'm impressed.

Out of the thousands of drives already sold very few, per volume are having issues. I'd say most are being used on Z68 and P67 chipsets so you will find that most people having issues are also using the same chipsets.
 
Bottom paragraph of page 4:

The Marvell SATA III bridge chips found on many X58 motherboards will now allow your 2011 SATA III SSD to reach peak read or write speeds.

Shouldn't this read "will not allow"?

Also on page 6, shouldn't the lowest entry in the graphs be the Agility 3? Looks like the label and values were carried over from the Vertex 3.

English shmenglish, I always rely on [H] for solid real-world reviews. Keep up the good work! :cool:
 
Last edited:
I had a 240GB Vertex 3 paired with a GA-P67A-UD7-B3. Obviously I was trying to use it with one of the Intel 6Gbps ports. The first firmware (2.02) wouldn't work with the Microsoft AHCI driver, so you could only get Windows 7 to install by preloading the Intel RST driver. How this got past any kind of testing before release to market is beyond me. After install Windows would still periodically blue screen. None of the newer firmware versions have resolved the problem so far, including the latest 2.09 (although this did seem more stable). I've been able to easily reproduce these problems with three different 240GB Vertex 3s (all first batch drives purchased early April) and three different P67 boards from different manufacturers.

OCZ support has been useless. They seem to be more interested in covering their ass than admitting the problem. Most of the advice they offer sounds like delaying tactics designed to waste people's time, such as endlessly reinstalling Windows multiple different ways. The rest of their advice sounds like faith healing that makes no sense (not that I haven't tried it anyway). There really doesn’t seem to be any logical reason why many of their suggestions would have any impact on anything related to the drives or drive controllers, or anything involving the entire disk subsystem. It almost sounds like they’re a bit desperate, despite assuring people that almost no one is having trouble. They might as well be telling people to wait until the next full moon when Venus and Mars are aligned, then slaughter a goat to the Vorlon god Boogee while throwing salt over your left shoulder, and if that doesn’t work, put the SSD in the freezer.

I've temporarily replaced the Vertex 3 with a Vertex 2 and I've had no issues at all. It feels so much better knowing I can just rely on my system to work now. I’m very happy that I've given up on the Vertex 3.
 
It was noticed. I sent Kyle the correct images and he will get them uploaded when he sees them. The numbers at the bottom are the Agility 3 240GB. I think the information on the next page is a bit more important though.
 
It was noticed. I sent Kyle the correct images and he will get them uploaded when he sees them. The numbers at the bottom are the Agility 3 240GB. I think the information on the next page is a bit more important though.

The numbers are for the Agility? But the numbers are exactly the same?
 
You're right, I briefly relooked at it. I already had corrected chart made but somewhere along the line I didn't update the back end files or I attempted to and didn't do it right. It's been years since I worked with a back end system and am still shaking out the dust.
 
I was lucky to be a day-1 adopter and got the original 2.02 firmware on my 240's. I haven't touched them as far as firmware.

The drives have been nothing but blazing fast and rock solid in RAID0. No driver issues, no BSODs, nothing weird at all, ever. Best SSDs I've had yet and I've had a ton of them. Everything I use boots / runs off the array, including all of my Steam games.

Pretty much the perfect storage solution for my usage. Only thing more perfect will be 2 more on an X79 board. :)
 
You're right, I briefly relooked at it. I already had corrected chart made but somewhere along the line I didn't update the back end files or I attempted to and didn't do it right. It's been years since I worked with a back end system and am still shaking out the dust.

I'm just wondering how I was the only regular reader to notice this. Am I the only reader who actually looks at the perty charts you put together for us!?!?!?
 
"Once you get past the form factor issue you have a SSD that costs half of what last year’s model did at launch and is capable of twice the performance."

Maybe next year I'll bite then. :D
 
The Marvell SATA III bridge chips found on many X58 motherboards will now allow your 2011 SATA III SSD to reach peak read or write speeds.

Obvious typo aside, rather than being very vague about the marvell controller not being as fast, how about an actual indication?

I only have the Marvell controller, and based on your benchmarks I would say its about a 10% penalty to the top transfer speed, and I'm using the BSOD workaround firmware which is apparently slower.

the way you worded it came across as "if you don't have native SATA 6GB, don't get this drive" but IMO 50MB/s less is still WAY FASTER than a Vertex 2.
 
Obvious typo aside, rather than being very vague about the marvell controller not being as fast, how about an actual indication?

I only have the Marvell controller, and based on your benchmarks I would say its about a 10% penalty to the top transfer speed, and I'm using the BSOD workaround firmware which is apparently slower.

the way you worded it came across as "if you don't have native SATA 6GB, don't get this drive" but IMO 50MB/s less is still WAY FASTER than a Vertex 2.

I agree. I see the same hit as you with the Marvell controller. But it's still faster than SATA 3Gb.
 
Consider yourselves lucky. Most x58 Marvell users are seeing reads in the mid 300's with even more dismal randoms. The first thing out of their mouths is "OCZ lies! The drive doesn't perform as advertised!" The problem is the Marvell controllers are crap as they've always been. I'm glad Kyle brought it up.
 
Last edited:
The unifying thread for people with BSOD issues with the SF-2281 drives is that most, if not all, are using the Intel 6GBps controller (P67/H67/Z68). I was one of them - I had an Agility 3 for about three weeks shortly after it first came out, and nothing I did made the BSOD problem go away. The tech support guy, who I was being nice to, basically told me something was wrong with my system, it wasn't the drive, and I should just return it to the retailer if I wasn't capable of figuring out what the problem was. When I politely informed him that I couldn't do that because the retailer didn't accept refunds, his words were, and I quote, "Not our problem. Ask them if they can make an exception." *click*

And all I did was tell him what the drive was doing and ask if I could RMA my drive or if they had any known fixes.

After OCZ basically told me to go sit on a shaft, I did manage to convince Newegg to let me return it for credit in spite of their no-refund policy on SSDs. I got an Intel 510 and *gasp* it worked without any dicking around.

Meanwhile, the complaints over on the OCZ forums all showed the same symptoms. OCZ went weeks saying it was not a problem with the drives, going so far as to differentiate their drives from Corsair's (who issued a recall for exactly the same problem). The growing multitudes that hung on went through multiple secure erase and new firmware cycles only to discover that not only did the BSOD/detection problems not really get sorted, now some drives are showing 50% (or more) performance degradation with the newest firmware. OCZ's reaction? It's your overclock. And your cables. And the fact that you're using your computer. And because you smell funny. Because it's certainly nothing to do with the drive, unless enough people complain.

To be fair, Corsair is having similar problems with the Force 3, even after the recall/rerelease and delayed the Force GT for the same reason. People are pissed off there too. But at least they're admitting that there's a problem and are sending out a mea culpa instead of telling everyone on the board to piss off.

I got burned once by OCZ around 2002-ish when they sent me and practically all of their customers RAM that didn't work at anything close to their specifications, and then magically disappeared when anyone tried to RMA anything. Remember that? IIRC Kyle and Anand both nearly got into legal fisticuffs with Ryan Petersen over that shit.

Now it's 2011, and I nearly got burned by them again. (A friend of mine literally got burned by an OCZ power supply that burst into flames last summer ... and which OCZ refused to RMA because they had forgotten to put a warranty sticker on it. But that's a different story.) To me, it's still the same shady-ass company as it was 10 years ago, just bigger and with more industry clout. They seem to only take care of their customers when the public eye is on them.

Anyway, I'm way off topic. Caveat emptor to anyone buying any SF-2281-based drives for Sandy Bridge platforms right now, OCZ or not. It's a roll of the dice as to whether your system will puke it up or run fine. Good news is that Sandforce is apparently aware of the problem and it sounds like more firmware iterations are forthcoming.

</rant>
Great post - I have been following the support forums and it's just absurd... I feel bad for the OCZ customers who have gotten the shaft. Its awesome that Newegg stepped up and let you swap for the Intel drive though.

I did a post on my blog about the OCZ quality issues and how they should issue a recall. Feel free to contribute any comments... the goal is to get OCZ to notice the glaring product quality issues.... which they seem content to ignore. Shipping SSDs that BSOD and don't even fit in a run-of-the-mill Dell notebook is simply unacceptable. I was thinking about trying an OCZ drive in my next PC, but not anymore.

http://techinsidr.com/is-a-recall-brewing-at-ocz/
 
Another question about form factor. Some of the slim laptops can only fit 7mm high laptop drives. Right now, that means spinner drives are limited to 320GB and ZERO SSDs fit without modification. Some drives have a spacer that can be removed, but that involves cracking open the case and voiding the warranty.

If the vertex III came in 7mm height I would have bought one already for my lenovo T420s. As it is, I went for a much much slower and smaller intel 310 mSATA drive because that's the biggest I could fit into the laptop without performing drive case surgery.

So how about it... 7mm height option for SSDs? Why not ship as 7mm with an optional spacer, instead of shipping as 9mm or larger, requiring warranty voiding to make it fit?
 
You're right, I briefly relooked at it. I already had corrected chart made but somewhere along the line I didn't update the back end files or I attempted to and didn't do it right. It's been years since I worked with a back end system and am still shaking out the dust.

So, was the Vertex 3 240Gb on there twice, or are the second numbers actually for the Agility 3 240Gb?

If the latter is true, the Agility 3 seems to be the drive to get, as it is cheaper and performs identically...
 
Zarathustra[H];1037442919 said:
So, was the Vertex 3 240Gb on there twice, or are the second numbers actually for the Agility 3 240Gb?

If the latter is true, the Agility 3 seems to be the drive to get, as it is cheaper and performs identically...

Actually, I just looked on Newegg...


The Vertex 3 240GB retail version is more expensive, but the OEM version is only $5 more than the Agility 3 240Gb retail version (The Agility 3 doesn't appear to come in an OEM version, at least not from Newegg)
 
Don't get the OEM version - it's not the same drive. Definite differences in performance from the retail V3.
 
Hmm.. I just put in an order for 2x240 and 1x120 ($1600 ouch :/) for a new server build on P67 platform, running server 2003 (i know... OLD lol) ... the blue screen thing is troubling :/
 
Hmm.. I just put in an order for 2x240 and 1x120 ($1600 ouch :/) for a new server build on P67 platform, running server 2003 (i know... OLD lol) ... the blue screen thing is troubling :/

Wow someone didn't do their research before blowing $1600 good luck!
 
Great post - I have been following the support forums and it's just absurd... I feel bad for the OCZ customers who have gotten the shaft. Its awesome that Newegg stepped up and let you swap for the Intel drive though.

I did a post on my blog about the OCZ quality issues and how they should issue a recall. Feel free to contribute any comments... the goal is to get OCZ to notice the glaring product quality issues.... which they seem content to ignore. Shipping SSDs that BSOD and don't even fit in a run-of-the-mill Dell notebook is simply unacceptable. I was thinking about trying an OCZ drive in my next PC, but not anymore.

http://techinsidr.com/is-a-recall-brewing-at-ocz/

I would never touch one of these drives if you paid me too.

An OS drive that BSOD's at random is unacceptable!!!

I'm glad the last product I bought from them was DDR 500 Modules!

I do feel bad for all of those stuck with bad drives, but this day an age if you don't research your products before buying them you are asking for this BS.

Everything these days is bug filled and the customer can beta test for us its so sad.
 
OK, sorry to take so long, been AFK. Corrections made to the pictures and the graphs on page 6.
 
I really think you guys are blowing this BSOD thing up. It isn't every drive and only a very small number of people are actually reporting the problem when compared to the number of people who've bought and started using them.

Believe me, if it was something that effect more people I'd be the first to write a front page feature on the story. I'm the only one to do anything about the Agility 3 performance issue when the drive has data on it and the only one to really report on the form factor issue. As as result I don't have a RevoD here to tell you about today.

To be honest I really think the form factor and Agility 3 issues are bigger issues since they affect everyone. Hang tight and I'll get the Corsair Force 3 and Patriot Wildfire articles ready for the first of next month. The Patriot Wildfire is a direct competitor to the Max IOPS and the Force 3 is a competitor to the Agility 3. Corsair says they do not have the performance issues I ran across with the Agility 3 and if it's true the price / performance will be amazing.

The Force 3 arrived an hour ago. I'm headed to the lab now:)
 
As as result I don't have a RevoD here to tell you about today.

Interesting. i guess they are a bit sensitive about the fact that their shiny new $1700 revo drives don't have TRIM support, eh?

I guess Anand got away with this by mentioning that there would be a performance issue, but not actually testing it and comparing it on a chart.

Besides, I guess he's become Mr. SSD now, so they pretty much have to have him review drives regardless.
 
I would never touch one of these drives if you paid me too.

An OS drive that BSOD's at random is unacceptable!!!

I'm glad the last product I bought from them was DDR 500 Modules!

I do feel bad for all of those stuck with bad drives, but this day an age if you don't research your products before buying them you are asking for this BS.

Everything these days is bug filled and the customer can beta test for us its so sad
.
You hit the nail on the head. OCZ is effectively outsourcing their validation department to their customers. For me personally, that is not acceptable for a high ticket product like an SSD.

I'll always buy other SSDs that work fine out of the box... even if they're slower.
 
Why don't you test them in actual real world application?

I don't care about how nice transfers those drives can get in canned benchmark (where Sand Force rocks due to data compression)

Show us how much time we can save when loading games/applications (os is done in article fortunatly) by paying price premium for Sata 3 drives over sata 2 cheaper ssd.
 
Why don't you test them in actual real world application?

I don't care about how nice transfers those drives can get in canned benchmark (where Sand Force rocks due to data compression)

Show us how much time we can save when loading games/applications (os is done in article fortunatly) by paying price premium for Sata 3 drives over sata 2 cheaper ssd.

Give us some more specific scenarios that you would find useful.
 
Back
Top