Comments Locked

107 Comments

Back to Article

  • JWatson - Wednesday, June 15, 2011 - link

    Did OCZ by chance fix the firmware issues with this drive as well as increase the speed? There are still countless posts of BSOD and crashing related to the Vertex 3.

    http://www.ocztechnologyforum.com/forum/showthread...
  • kensiko - Thursday, June 16, 2011 - link

    FW 2.08 was posted yesterday. It did help a few people but there is still some persons with problems.

    Last night, SandForce was able for the first time to reproduce the BSOD issue. Expect a new firmware once it's validated.

    It seems they are working day and night.
  • 7Enigma - Thursday, June 16, 2011 - link

    They might be working day and night, but I would be hard-pressed to purchase or recommend any Sandforce-based drive until this is all sorted out (both Vertex 2 and Vertex 3 issues).

    I really liked Sandforce and OCZ when they debuted the Vertex 2, but they have seriously damaged their reputation between SSD's and their PSU's.

    I just hope Anandtech takes a harsher stance on them in the coming reviews as recently it appears more of an afterthought in passing comments than a real criticism.
  • noeldillabough - Thursday, June 16, 2011 - link

    I'm sure glad I went with the Intel 510, sure its not as fast as the Vertex but I have had NO problems whatsoever with the drive since it was launched. (using it in a P170 Clevo notebook)
  • semo - Friday, June 17, 2011 - link

    I agree. Review sites in general are being very nice to OCZ. Even Storage Review have backed off the 25nm transition fiasco (even though there has been no official recall issued by OCZ).

    What is being discussed in forums is not reflected by review sites at all. Everyone is bashing OCZ, while the company's products are being awarded left, right and center. This is becoming very much like politics unfortunately. The people we look up to are doing things that seem irrational to us because they have vested interests and fear "burning bridges".
  • L. - Friday, June 17, 2011 - link

    I believe its more of a misunderstanding :

    OCZ Vertex 3 has wonderful specs, and provided it doesn't BSOD on the reviewer, has all the reasons to rank very high.

    On the other hand, quite a few people seem to have reliability issues, but how do you expect that to impact a review ?

    Important fact too, the message conveyed here on Anandtech is that Intel drives are more reliable - it may very well represent the Intel vs OCZ comparison taking into account your comment as an unhappy OCZ customer.
  • semo - Friday, June 17, 2011 - link

    In the original article, Anand is only praising the Vertex 3 MAX IOPS. He already knows about the BSOD issues so why not include that in there as well as a disclaimer (this drive could be great, but be cautious , etc...).

    I also think that reviews should be updated if the reviewed product is no longer relevant. The original Vertex 2 was great but if you decide to buy one from ebay, you have no chance of knowing what crap you might get because the model name is worthless for describing the drive. Even the spec sheets for the Vertex 2 can not be trusted at all.

    Do you see Intel changing the hardware inside their chips without changing the model numbers and deceiving customers? Have you ever bought a hard drive from WD only to find that the capacity is different to what is written on the box/spec sheet/review? This sort of stuff is scandalous but OCZ has walked away from it all unscathed.
  • Tros - Sunday, June 19, 2011 - link

    "Do you see Intel changing the hardware inside their chips without changing the model numbers and deceiving customers?"
    I recall an issue where customers would buy quadcores that advertised VT-x, but found it was an utter-mix-up in terms of which processor batch (indistinguishable on purchase) you got as to VT-x or not.

    "Have you ever bought a hard drive from WD only to find that the capacity is different to what is written on the box/spec sheet/review?"
    There's the whole 1024 versus 1000 thing. I'm not sure if you're being ironic or ignorant.

    "This sort of stuff is scandalous but OCZ has walked away from it all unscathed."

    No, their reputation is still questionable by the comments, and even by the writing on Anandtech. But the case in the world is that a few people get lemons, and have the need to drown out everybody else who did just fine. No lemons here, and it's been nearly a year and a half with my vertex 2.
  • semo - Sunday, June 19, 2011 - link

    Link to the VT-x issue?

    There is no issue about binary k and decimal k. Both are valid. If you buy a 1TB Hitachi and WD, they will both be the same in capacity (or very close to). If you buy a 1TB Hitachi again 6 months down the line, then it will be the same capacity as the first one. Sounds logical but that's not how it works with a "60GB" Vertex 2.

    I don't think you know the details of the 25nm transition fiasco. A year ago OCZ drives were great but they started changing the specs dramatically around December 2010 in the retail channel
  • Tros - Tuesday, June 21, 2011 - link

    I am not familiar with that issue, and am looking it up right now. I did just get a 120 GB drive, since my 60 GB one is wonderful.

    http://www.slashgear.com/intel-cpu-virtualization-...

    ^ Not what I thought it was. Same ballpark, but remedied by looking up data sheets on the processor you want at the time.

    *: I looked up the issue, but after skipping a 1TB drive (because it'd never display 1TB under Windows formatting), I'm really okay with this.
  • Anand Lal Shimpi - Saturday, June 18, 2011 - link

    I have tested 8 different SF-2281 based drives and unfortunately I haven't been able to get any of them to create the BSOD issue that other users are seeing. I have also deployed three OCZ SF-2281 drives full time in machines here (one on 24/7) with hopes of accelerating the problem, but again, I haven't seen a single issue. Ever since my first SSD review I've done extensive long term testing on any drive we recommend to get a good idea of its overall behavior over time and reliability. Regardless of whether others were having issues this is standard procedure for us here. If we recommend it, we have to use it in our systems.

    That being said, I know the problems exist, I just don't know why and I also don't know how widespread. At this point any SF-2281 recommendation is basically "hey this could work, but keep in mind that you could have problems. if you don't want the risk, go Intel".

    As always, I'll keep on it. I spent three months testing the Intel SSD 510 in my main system before swapping it out for a Vertex 3. I'll be testing it for the next few months as well to see how it behaves over time.

    Take care,
    Anand
  • Jeff72 - Saturday, June 18, 2011 - link

    If you want to get the blue screen:

    I used an Asus P8Z68-V PRO motherboard with current BIOS (I was using the beta 8801 BIOS at that time) at stock (no overclocking) settings
    CPU was Intel i5-2500K at default stock settings
    Graphics card is AMD 6950 at stock settings
    OCZ Vertex 3 120GB (with firmware 2.06 at the time) (for OS)
    Western Digital 640GB 7200RPM hard drive (for data)

    Set SSD in BIOS to AHCI and Hot Plugable, per OCZ recommendation.
    Updated SSD from 2.02 to 2.06 firmware.

    Partition SSD before installing Windows 7 (Boot from Windows 7 DVD and use Shift-F10 and "diskpart" to partition for one partition)

    Installed these drivers only:
    Intel LAN Driver
    Intel Chipset Driver
    Realtek Audio Driver
    Management Engine Interface (Intel)
    USB 3.0 Driver
    ASUS Bluetooth Driver
    Graphics card drivers

    Installed these applications:
    Firefox
    Chrome

    Ran Windows Experience Score to Rate this computer

    Leave power settings on Windows 7 at default (allow it to sleep).

    Leave the system on and at times sleep on occasion during the day and browse the web with two Chrome windows open. Eventually I got the blue screens on two separate brand new OCZ Vertex 3 120GB SSD drives after installing as above. The system would at times slow down and seem to lock up and eventually blue screen.

    The Windows 7 seemed to blue screen because the SSD OS drive seems to vanish. After Windows 7 warm reboots the system from the blue screen, BIOS does not see the OCZ SSD drive, but it did still see my data hard drive in BIOS. After powering off and powering on the system completely, the OCZ SSD was seen in BIOS again. Sometimes I would not get blue screens for awhile, other times it happens fairly quickly.

    Same system and install but now with an Intel 510 series 120GB SSD drive and I have no issues now. No blue screens at all anymore.

    Note, I did NOT install Intel RST at all yet on any installs since I do not believe I needed it since i'm not using RAID.
  • rjraouf - Wednesday, September 14, 2011 - link

    Hi,
    I have the same system configuration as Jeff72, Asus MB, Vertex SSD and OS. I have been having the same issue. OCZ recommended upgrade to latest firmware 2.11 and secure erase, and fresh install of OS. Problem continues to re-occur.

    OCZ is not a company to be trusted and they DO NOT stand by their product. They even refuse to acknowledge a problem.

    I can pretty much guarantee if anyone wants to replicate the problem, simply get an Asus P8Z68-V PRO motherboard and attach the SSD to either the Marvell or Intel controller and after installation of Windows 64bit, Run some applications and then let the computer sit ideal for a while. About 12-24hrs from that point, the symptoms will appear. Computer reboots, SSD is not recognized unless you shut the power off and on. If you have trouble making it reboot after 12hr of ideal time browse the internet for a few min and either reboot occurs while browsing or after idling the system again.

    I have replicated this problem successfully three times on three sets of motherboards and SSD drive
  • semo - Sunday, June 19, 2011 - link

    Thank you for this info Anand. Why do we have to read through the comments section of every SSD article to find useful nuggets of information like this? If you know about an issue in the wild, why not write a news article about it?
  • Holly - Tuesday, June 21, 2011 - link

    Asus P8P67 PRO (latest non-beta bios, board revision 3.0)
    OCZ Vertex 3 240GB (2.06 firmware)
    2x WD Black 640GB (raided data storage)
    1x WD Raptor 150GB (system backup)

    Intel Rapid Storage Technology driver version: 10.5.0.1027

    Problems with drive started about 24hrs after being installed (along with clean Win7 64b installation)

    To make system even remotely stable I had to perform all the steps mentioned in http://geekmontage.com/texts/ocz-vertex-3-freezes-... (well, except turning off raid mode in favor of ahci mode which I could not do)
    Then I had Vertex causing BSODs and not showing up after computer restarted (another obviously wide spread problem).
    It seems finally there might be light at the end of tunnel after updating firmware to 2.08, but it's too soon to tell... Among other things updating firmware on your system SSD is definitely not a thing regular computer user should even think about.

    The drive is helluva fast, I like it, but it would be fine if I could actualy use it without bsod worries. Especialy if I want to use remote desktop.
  • Jeff72 - Thursday, June 16, 2011 - link

    I bought an OCZ Vertex 3 120GB SSD. Installed Windows 7 64bit and the needed drivers for my Sandybridge Asus P8P67-V PRO motherboard. Eventually got blue screens in Windows and it would reboot itself and then BIOS would not see the SSD anymore until a full power off and power on was done. I exchanged the SSD for a replacement. Installed Windows 7 64bit on 2nd exchanged OCZ Vertex 3 120GB SSD and the needed drivers for my Sandybridge Asus P8P67-V PRO motherboard. Eventually got blue screens in Windows and it would reboot itself and then BIOS would not see the SSD anymore until a full power off and power on was done. I returned the 2nd OCZ Vertex 3 120GB SSD for a full refund.

    I bought an Intel 510 series 120GB SSD. It just works.
  • Jeff72 - Thursday, June 16, 2011 - link

    Correction: Sandybridge Asus P8Z68-V PRO motherboard.

    The Intel SSD is working great and no blue screens at all. Very fast and reliable is a good thing.
  • L. - Friday, June 17, 2011 - link

    Makes me wonder how good it works on non-Intel motherboards tbh.
    I wouldn't put it past Intel to create issues between their chipsets and other people's SSD's.
  • rpmrg - Friday, June 17, 2011 - link

    They work fabulous on AMD's SB850 SATA 3.0 controller.
  • Belard - Friday, June 17, 2011 - link

    I've installed intel SSDs into AMD systems... no problems. Other than I don't like the silly AHCI screen that pops up during POST for a second. Otherwise, no issues.

    With all the reviews I've seen... OCZ Vertex3 seems like a good choice, in general, but the intels still look pretty good and their reliability record still seems to be there.
  • Tros - Sunday, June 19, 2011 - link

    Seriously.

    I recall the lawsuit for their compiler specifically disabling optimizations if a non-Intel chip was running the code: http://techreport.com/discussions/8547.
  • aldiswilson - Saturday, June 25, 2011 - link


    ONLINE STORE :
    www styl eown com
    -------------------------------------------------------------------------------------------------
    1) More pictures available on our website --
    3) Perfect quality, small order accepted .
    4) 100% safe door to door delivery, within 5 - 7 days air express for small orders .
    5) We have lots of jerseys in stock
    -------------------------------------------------------------------------------------------------
    www style own com
    -------------------------------------------------------------------------------------------------
    6) Letters and number are sewn on b2cshop body, 100% embroidery
    7) Size: .48, 50, 52, 54, 56, 60
    8) Delivery by UPS, DHL, EMS door to door
    9) Delivery in 5 - 7 days
    -------------------------------------------------------------------------------------------------
    NFL,NBA,MLB all are 18usd!!!!
  • aalbionthomas - Sunday, June 26, 2011 - link


    ONLINE STORE :
    www styl eown com
    -------------------------------------------------------------------------------------------------
    1) More pictures available on our website --
    3) Perfect quality, small order accepted .
    4) 100% safe door to door delivery, within 5 - 7 days air express for small orders .
    5) We have lots of jerseys in stock
    -------------------------------------------------------------------------------------------------
    www styl eown com
    -------------------------------------------------------------------------------------------------
    6) Letters and number are sewn on b2cshop body, 100% embroidery
    7) Size: .48, 50, 52, 54, 56, 60
    8) Delivery by UPS, DHL, EMS door to door
    9) Delivery in 5 - 7 days
    -------------------------------------------------------------------------------------------------
    NFL,NBA,MLB all are 18usd!!!!
  • Warlock15th - Wednesday, June 15, 2011 - link

    It's good these drives finally get to be tested, I've had a couple of these in RAID 0 since the day they were available, coupled to an Asus MIVE and a Ci7 2600K in SATA III, and the performance is just amazing; yes there were some problems when the drives came out, but if you use the latest IRST drivers from Intel and update your drives to the latest firmware you won't believe how fast these drives can be.

    I'm really looking forward to the review.
  • Stuka87 - Thursday, June 16, 2011 - link

    And by SATA III we can assume you mean SATA 6.0? Since SATA III does not mean anything according to the Serial ATA International Organization. It is actually explicitly stated to never use that naming scheme because it is made up and means nothing.

    Sorry, but its kind of a pet peeve when people use false terminology.
  • Drag0nFire - Thursday, June 16, 2011 - link

    I see. Then this website from the Serial ATA International Organization describing "SATA revision 3.0" must be mistaken.

    http://www.serialata.org/technology/6Gbdetails.asp

    You should probably tell them about your pet peeve. Maybe they will change it for you.
  • Warlock15th - Thursday, June 16, 2011 - link

    SATA Revision 3.0 (III stands for 3 for those illiterate with roman numerals) is the latest SATA spec protocol that allows for up to a 6GBps max bandwidth for data throughput for all current drives that use said protocol.

    Don't know what time in the future you come from, but we here in 2011 still have to contend with SATA 3.0 speed limitations when it comes to SSD drives performance, not that any current existing drives can max the available bandwidth as of June, 2011, can you please enlighten me as to how fast we can expect this hypothetical SATA 6.0 protocol to be a few years from now?
  • Warlock15th - Thursday, June 16, 2011 - link

    And yes, I know it's "Gbps", not "GBps" as I posted... That little "b" there makes for a world of difference.

    It seems I've also joined the "visitors from the future" bandwagon, by citing a spec that's an order of magnitude faster than what's currently available :p
  • twhittet - Thursday, June 16, 2011 - link

    Yes, I don't know this SATA 6.0 either. A pet peeve of mine is when other people correct someone, and are also wrong.

    Sorry, but its kind of a pet peeve when people use false terminology.
  • KaRRiLLioN - Thursday, June 16, 2011 - link

    I think he means SATA Three. I don't speak Roman but after someone clarified the Roman Numeral thing, then it made total sense to me.
  • L. - Friday, June 17, 2011 - link

    Errrr go to school, it's great ?
  • seapeople - Monday, June 20, 2011 - link

    He must get really confused when watching the Superbowl.
  • Movieman420 - Wednesday, June 15, 2011 - link

    Ocz are working feverishly with SandForce as well as affected users in an effort to reproduce this issue on a test rig which can then be analyzed with a good chance of finding the culprit. This issue is also present in OWC and Corsair drives as well. Tho there are several bsod posts in the support forum but according to the support team this issue affects less than 1 percent of the new SF2200 drives. Keep in mind that you rarely ever see positive posts on a support forum. Anyone having the issue can probably get a refund if they don't want to wait for the fix. This is a compatibility issue as it occurs across many different chipsets (amd and intel). Numerous RMAd drives have been tested by Ocz and SF and all have tested with ZERO issues and those who received a replacement drive still have the same issue. There's a single common thread that these affected machines all share, be it related to bios settings, chipset drivers or a million other possibilities...once they can replicate it under laboratory conditions they should be able to chase it down. Even tho the issue affects only a verrry small percentage of users, that doesn't make it any less frustrating for the folks who shelled out $300 to $500 for the lastest and greatest that the ssd world has to offer.
  • TrackSmart - Wednesday, June 15, 2011 - link

    Yes, they are working on it, like the previous SandForce SSDs... but don't expect a refund if it is not compatible with your system. OCZ and Corsair will not help you.

    To be clear, I don't have a problem with the fact that there can be issues with new SSDs. My problem is how the SSD makers (OCZ and Corsair in particular) have dealt with it, which is to blame SandForce for the problem and then tell customers they have to wait for new firmware. No refunds or exchanges for a different model are allowed. I hate when the customer is left holding the bag for a problem that is beyond their control.

    If only this message could be passed to the higher-ups at OCZ and Corsair. Let them know they are losing formerly loyal customers, like myself. That's a bad strategy in the long run.
  • kensiko - Thursday, June 16, 2011 - link

    They had refund some people. Ask them if you want.
  • dubyadubya - Wednesday, June 15, 2011 - link

    I have a Vertex 3 MI drive and the BSOD issue. I have had the drive about 45 days and have had 4 BSOD. Days or weeks can go by with out a problem. I had thought a workaround had fixed the problem as the PC had been trouble free for 15 days. Just today BSOD. Keep in mind this same PC ran for 9 months without a single BSOD using an Intel X 25-M G2. When the problem happens windows 7 freezes followed by a BSOD STOP: 0x000000F4. The system is usually at idle or near idle when the problem occurs never under load. Once this error occurs the drive is no longer detected by the bios on post and a cold boot is required to fix the problem. I have tried every possible fix OCZ recommends but no dice. I can see how hard it would be for OCZ/Sandforce to replicate the problem considering it could take days or weeks to happen just once. There are no errors logged because the drive is off line so nothing can be written.
  • kensiko - Thursday, June 16, 2011 - link

    Try firmware 2.08
  • MrSpadge - Thursday, June 16, 2011 - link

    Well, there's an easy fix for that then: BOINC can keep your Pc under full load quite reliably ;)

    MrS
  • kavanoz - Thursday, June 16, 2011 - link

    Same here, I replaced my X25-M G2 with Vertex 3 MI. But instead of days between the BSODs, I get them daily. I leave my home system on 24/7 as I remotely work on it from my office. I have to turn my PSU off and on everyday when I get back home. Following up OCZ forums for the permanent fix but I don't know how much longer will I be able to wait.
  • geddarkstorm - Thursday, June 16, 2011 - link

    Has anyone thought it could be a Windows 7 "power saving" option that's the culprit? That whole "turn off harddrives" after x amount of idle time? Maybe the OS, with whatever Windows updates you've installed, is glitching when it comes to that (for instance, hibernate and sleep are both very glitch prone things with Windows on any hardware in my experience; just can't trust it to manage power settings correctly on a PC).

    I wonder if you change your power management advanced settings if the issue will stop. Power management is definitely a common thread, since this is happening at idle, when that would be invoked.
  • kavanoz - Thursday, June 16, 2011 - link

    As recommended by OCZ, sleep and hibernate and all hard drive power saving features are disabled (remember I need the PC on 24/7 so i can access it remotely anytime), hot-plug enabled, Intel RST drivers removed, other controllers tried (Marvel), Windows re-installed etc. etc.

    BTW, I have most of the BSODs when the computer was idle, but there was one time when it happened while I was actively using it. Just BSOD in front of my eyes.

    They said Sandforce re-produced the problem in the lab, so hopefully they nail this down soon.

    http://www.ocztechnologyforum.com/forum/showthread...
  • seapeople - Monday, June 20, 2011 - link

    Does the extra speed of your blue-screening Vertex drive make you exceptionally happy?
  • chrnochime - Friday, June 17, 2011 - link

    Refund because it's incompatible with your PC? And you're saying this with a straight face?
  • OWC Grant - Monday, June 20, 2011 - link

    It has NOT been reported to us, nor have we found any forums that discuss OWC SSDs creating the BSOD like other SSD brands, noteably OCZ and Crucial, are creating.

    If you have such a forum address, please let us know that. Otherwise, pls make sure you are fully informed as to the reliable status of OWC SSDs.

    Thanks!
  • Beenthere - Wednesday, June 15, 2011 - link

    Corsair has had to recall their latest SSDs days after release. OCZ hasn't recalled their bad drives yet and people are furious as they well should be. As usual many vendors are rushing this half-baked tech out the door for record profits at the expense of consumers. Microsucks pioneered this ship it and fix it later mentality and it's worked well for them so others have followed. It's criminal IMO and the vendors who do this should pay treble damages and this crap would stop in a heatbeat.

    I am interested in SSDs for laptop use and have looked at some tests of identical laptops other than one had a 500 GB 7200 rpm HD and the other a 120 GB SSD. It was interesting to see that in the case of the Dell laptops both scored the same performance rating. In two other brands the numbers varied about 10% better in one case for the HD vs the SSD. So it was a split decision. I wonder if typical laptop CPUs limit the advantages of a SSD or what exactly is different in a laptop vs. a desktop?????? Curious minds want to know as it would seem that a 2.5" laptop HD would be a real bottleneck even with a 500 GB 7200 rpm HD compared to a SSD.
  • TrackSmart - Wednesday, June 15, 2011 - link

    I have SSDs in desktop and laptop computers. Except for sequential reads and writes (where spinning hard disks still do pretty well), there is no question of greater overall performance for most SSDs. I don't know what kind of benchmarks you were running.

    But I agree that there are lots of compatibility issues between SSDs and laptop hardware. Problems that are exceedingly rare for ordinary spinning hard disks. With my desktop, and I had great luck with the first SSD I purchased. For my current laptop, it took three tries to find a model that didn't have problems with sleep/hibernate/boot-up. The common link was SandForce based drives. A Crucial C300 is now in that laptop and working flawlessly, though others have had issues with that drive and some laptop hardware combinations. I guess it will be a while before you can expect an SSD to 'just work' in a given system, as we've taken for granted with spinning hard disks.

    Now if only the SSD makers would publish some data on what combinations of hardware are known to be problematic. And offer better options for those who are suffering from compatibility issues.
  • Beenthere - Thursday, June 16, 2011 - link

    At this point I don't think the SSD folks fully understand the problem with the SandForce drives and what hardware is affected. It's a work in progress and consumers are the guinea pigs... as usual. Pay for bleeding edge hardware and you will bleed it appears.

    As far as the laptop performance I'm not the one who conducted the tests. I was just reporting on the results that were observed in a test of three different laptop models by different companies. I would love to see more back-to-back comparisons to see if the CPU or something else limits the full performance of an SSD in laptops.
  • 7Enigma - Thursday, June 16, 2011 - link

    Sorry man, whatever testing you are referring to was faulty (due to user setup error/measurement), or outright lies.

    The difference in general computing, ESPECIALLY on a slow/older laptop drive is downright amazing. Anandtech has done these over the last 2 years or so with their review hardware where they review a system and then pop out the crap HDD and replace with a SSD. The last one I can remember was a MacBook I believe but they have done it with several computers.

    Even on netbooks where the CPU can definitely hamper performance you see a benefit (though the cost typically doesn't make sense in this sector).
  • Beenthere - Thursday, June 16, 2011 - link

    I don't believe the testing was faulty. I believe the testing software used and the systems being tested have a significant impact on the results. Many reviews use only synthetic benches which measure block size transfers. This does not necessarily represent actual workloads. More realistic testing such as productivity, HTPC disk capture and gaming benches can often provide a clearer performance picture.

    Just to be clear I truly believe that SSDs should provide a substantial increase in performance especially in laptops but I want to see those results from real world testing not just synthetic benches which often look great but are not representative of actual PC performance. It would be interesting to compare the same SSD in a top of the line PC vs. a mainstrean laptop using real world benches to see how much the CPU, SATA II vs. SATA III etc. impact real world performance.
  • geddarkstorm - Thursday, June 16, 2011 - link

    Um... there have been tons of real world tests for SSDs verses HDDs in laptops and PCs over the years. Just look at boot time alone, for instance. No HDD will come close to an SSD in the speed of booting or application loading. I have no idea what you are talking about, but it is utterly false.

    See this http://eshop.macsales.com/Video/SSD/Performance_Te... for almost a DOZEN videos of SSDs verses HDDs in real world computing performance.
  • Beenthere - Thursday, June 16, 2011 - link

    I think you've missed the point... I'm not arguing the value of SSDs per se. I'm saying I would like to see real world benching of SSDs in laptops to see if the SSDs performance varies significantly compared to a typical HD and also in comparison to a desktop PC with more CPU power, etc.
  • geddarkstorm - Thursday, June 16, 2011 - link

    Those benchmarks are everywhere, all over the net. I just linked you to some that you can watch first hand, instead of just reading as numbers. Not many sites have collated it all into one easy package, but Google is your friend. You can find a plethora of real world benchmarkings (boot times, app loading, productivity, battery life, etc), of different SSDs on the same hardware and the same SSD on a huge spread of hardware, and then just compare.

    Every review site out there uses a different combination of hardware, so just start browsing more than one site. Try techreport.com , pcper.com, extremetech.com , techgage.com , thessdreview.com , and so so so many more. Everything you're looking for is already out there, just look.
  • Beenthere - Thursday, June 16, 2011 - link

    There in lies the problem... There are lots of different tests and little data of actual value. I've looked. This ain't my first day on the job. <LOL>

    That's why I pointed out the inconsistent test results and why I would like to see what I stated above done by Anandtech or some other reputable site. So far I have not seen this comparison.
  • Warlock15th - Thursday, June 16, 2011 - link

    I'm really sorry to hear people are experiencing this problem with these drives, as I mentioned in my post, at first I had some problems, specifically with Win7 randomly freezing from time to time, but after I updated my IRST drivers and the firmware, this behavior went away.

    I did post in the OCZ forums asking for help, and the community there was very assistive, I hope you guys can also solve your problems soon, as these drives rock, even more so when used in RAID 0 :)
  • jonup - Thursday, June 16, 2011 - link

    Bud, Anand has said it before, and I second on that, once you go SSD you never go back. It like with many things in life - you do not realize it until it's gone. I have a 3-4 year old lappy with Intel G1 SSD, it is the most responsive laptop in my office, it feels faster than SNB laptops with 7200 rpm HDD. I only have one desktop with 4xRAID0 HDD left and that is the least responsive PC that I own.
  • MrSpadge - Thursday, June 16, 2011 - link

    You do realize that Corsair only specifically recalled the 120 GB model? This is different from the BSOD issue.

    MrS
  • e-drood - Thursday, June 16, 2011 - link

    Presently OCZ, Corsair, OWC, Kingston are all waiting for Sandforce to analyze error reports/user complaints to base revised firmware update -- there may also be marchitectural design issues with Sandforce Series 2000 controller chips.

    While I can understand the Intel X58 chipset platform potentially having interface issues re:ssd/sata3; I absolutely do not accept P67/Z68 as having interface timing errors/voltage transients re: ssd/sata3. The Sandforce Series 2000 Controller is experiencing too many user complaints with too many different motherboards...

    Firmware alone will not resolve these incompatibility issues, neither will elaborate and arcane installation procedures.

    And I genuinely want to purchase additional ssd's as soon as possible...
  • semo - Friday, June 17, 2011 - link

    OCZ are a very anti consumer company. They still haven't issued a recall for their bad 25nm drives. People that haven't heard of the 25nm transition fiasco might not even know they have a dud.
  • spunlex - Wednesday, June 15, 2011 - link

    Is there a link missing at the end of the first paragraph?
  • Beenthere - Wednesday, June 15, 2011 - link

    Yes, the one below. Also read the article on "interleaving" which breaks it down in more detail.

    http://www.anandtech.com/show/4421/the-2011-midran...
  • shadowofthesun - Wednesday, June 15, 2011 - link

    Damn... I picked up the vanilla 120G Vertex 3 for my new build instead of the MAX IOPS edition about a month ago. I wish I had seen this first- I had thought the MAX IOPS edition was just a firmware tweak.
  • Zap - Thursday, June 16, 2011 - link

    "I ordered the 120GB MAX IOPS drive at the beginning of the week and just got it in yesterday"

    Anand, does this mean OCZ wouldn't send you one for review?
  • Griswold - Thursday, June 16, 2011 - link

    Companies like OCZ need to get through a full year without firmware horror and reliability issues before they can be considered by people who do more than just play benchmarks.

    So far, only Intel and Samsung seem to be able to deliver that. The rest is just good for benchmarking...
  • Beenthere - Thursday, June 16, 2011 - link

    I agree if a SSD isn't highly reliable then it is not welcome in any of my PCs. I'm not interested in constant firmware, driver updates and BSOD. Don't ship the product until it's sorted out properly. An SSD is worthless to me if it is not fully reliable.
  • L. - Friday, June 17, 2011 - link

    99% reliable might justify the price/performance ratio . Don't forget it's not 'all drives fail in all conditions' . it's more like 1% of drives fail in almost known conditions.

    Meh ;. if software was that reliable I'd be happy tbh.
  • 7Enigma - Thursday, June 16, 2011 - link

    That is definitely my opinion as well. My friend recently (last couple months) purchased a SSD for his Sager laptop and ended up going with an Intel G2 160GB drive. At the time the Vertex 2 and C300 were the kings of performance, but of very questionable reliability. He chose the Intel specifically because it was a validated product.
  • Movieman420 - Thursday, June 16, 2011 - link

    From Ocz website:

    'Unlike other brands OCZ does not use a reference design, and we design and manufacture our SSDs in-house, and are NOT affected by the hardware problems which are unique to other drive vendors.'

    They use their own proprietary design...Corsair uses a SF reference design, or is trying to.
  • Beenthere - Thursday, June 16, 2011 - link

    Agreed that OCZ is not using the reference design but they definitely have firmware issues. That's why I said that there are many unacceptable issues with SSDs at this time. IMO it's unscrupulous to ship these products without being 100% certain they perform as advertised and are reliable.
  • geddarkstorm - Thursday, June 16, 2011 - link

    No piece of hardware is 100% reliable, hate to say it. A 1% problem rate is really high, but who is to say it is their hardware alone that is the problem and not the operating system. Even Windows 7 is optimized for spinning platter based disks, and how to manage their power and work load, not for an SSD. All it takes is one bad Windows update (or combination of) or a corrupt driver to throw everything out of whack. If you've ever played with Linux and all the hardware fun that involves, you should understand what I'm saying.
  • Beenthere - Thursday, June 16, 2011 - link

    They have been able to duplicate the issues at OCZ, Corsair and SandForce. They just haven't identified all the conditions and stimuli for the operational issues. Operational issues are one matter but reliability is a whole nother matter. If an SSD doesn't perform properly or isn't reliable it's of no value to me and is in fact an expensive headache.
  • geddarkstorm - Thursday, June 16, 2011 - link

    Obviously that's true, for any piece of hardware, be it an SSD or an Intel SB motherboard.

    But this isn't an issue of reliability, the drives aren't physically breaking down, are they? Go peruse some HDD support forums and then compare reliability of SSDs to spinning platters.

    This isn't a perform properly issue either, not for the physical device. It meets all its specs. The only case I've really heard of where a modern SSD did not perform properly was when the defective 25nm flash was leaking into OCZ's Vertex 2 drives.

    Again, this isn't a reliability or performance issue of the physical drive. Something funky is happening with the OS/Bios and likely the power saving system. That's compatibility, but is it the SSD makers fault, or the fault of some bad OS update, bios/motherboard firmware? I guess we'll find out!
  • semo - Friday, June 17, 2011 - link

    How is this not a reliability issue if the drive BSODs the system? Take Windows out of the argument because after the BSOD, the SF SSD is no longer detected. I wouldn't trust the 1% figure too much. I just don't see it as a credible figure.

    Keep in mind also, that OCZ censor their forums and delete some threads that highlight the issue.
  • chrnochime - Friday, June 17, 2011 - link

    Just because they aren't physically broken doesn't mean they're reliable. That's like sayiing a bricked Router is reliable because all of its SMTs are still there and the flashchip that holds the firmware is still sitting on the board nice and soldered. It's not broken physically but then it doesn't do squat either.
  • kensiko - Thursday, June 16, 2011 - link

    I'm sure they were 100% their SSD was completely safe, but sure they did not test every hardware combination in the world. What would be the hardware combination ?
    100000000000000000000000000000000000
    ?

    No, 100000000000000000000000000000000000000000000000000000000000000000

    :P
  • L. - Friday, June 17, 2011 - link

    And so, man invented scientific notation, and he saw that it was good.
  • HaydenOscar - Thursday, June 16, 2011 - link

    I'm looking to get an SSD (~120GB) in the near future, and was pretty sure I was going to get an SSD 510. Vertex 3 may be faster in most cases, but the loss of performance made me feel like Intel's offer was better as it offered steady performance regardless of compressibility. However, the performance of this drive is just a bit too much for me to ignore.

    My question to you people is: is my uneasiness unfounded?
  • kensiko - Thursday, June 16, 2011 - link

    If you can get it at a good price than great, but don't forget this is a Marvell controller.

    In my situation Intel drives were rarely competitive (Canada). I got 2 X25-V when they were but now they are beaten by the others.
  • HaydenOscar - Thursday, June 16, 2011 - link

    I'm in Australia: standard edition Vertex 3 120GB is $299. Max IOPS edition is $329. SSD 510 is $369.
  • e-drood - Thursday, June 16, 2011 - link

    OCZ, Corsair, OWC, Kingston and others only resell ssd's based on Sandforce Series 2000 controllers - said ssd's are manufactured by small number of specialist odm's...

    None of the Resellers are Qualified To Determine Solution Of Problem(s) Being Experienced By SSD Purchasers/Users...

    And Sandforce is economically motivated NOT TO RECALL Series 2000 Controllers...
  • HaydenOscar - Thursday, June 16, 2011 - link

    I should clarify that it's not any of the "problems" experienced by users of these controllers I'm uneasy about, but rather that I don't want all my performance to be dependent on the compressibility of my data.
  • L. - Friday, June 17, 2011 - link

    IF you have a doubt, go ahead and determine the compressibility of your data :

    Take all you want to store on the ssd, right click add to winrar archive, best level of compression and then you'll have an idea.

    I think you are clearly underestimating the amount of uncompressed data that is normally stored on an ssd (os, programs, games).
  • Jeff72 - Thursday, June 16, 2011 - link

    If you buy an OCZ SSD, I strongly recommend you buy it form somewhere local or from a vendor with excellent return policy in case you get blue screens.

    Intel SSDs are just more reliable.
  • L. - Friday, June 17, 2011 - link

    Unfounded if it's based on that.
    The truth is EVERYTHING should be compressed, or nothing . or ... whatever

    The reality is that most stuff is not compressed, most game installs are not compressed (hey it took one dvd now it spans 16 gigs .. what do you guess :p) etc.

    You're not going to use that ssd for fully compressed stuff like video, photo or music, so you will enjoy a great deal of compression acceleration.
  • Belard - Friday, June 17, 2011 - link

    Are you the real L?
  • lenghui - Thursday, June 16, 2011 - link

    Anand, Do you have any articles about running VM on SSD? I am thinking in terms of performance, life span, and pros vs cons. I know there are many aspects to consider (Windows Hyper-V vs VMWare vs Xen). There are some discussions out there, but most are theoretical and based on personal opinion. I figure why not ask the expert (AT) and its expert readers. Many thanks.
  • L. - Friday, June 17, 2011 - link

    SSD excellent for VM's because of the IOPS/G ratio, however there are many known issues in hypervisors regarding disk io -> the bigger part of your solution is in software, not hardware, even though SSD's can handle many times more VM's per disk than HDD's.
  • semo - Friday, June 17, 2011 - link

    I've been looking for more VM coverage as well but it seems that it is not a high priority for most review sites. VMs benefit a lot from high IOPS but the general consensus amongst the public and critics is that "we have enough IOPS". Obviously I don't agree but if you look at the success of the Intel 510 (lower IOPS than predecessor) it seems that the industry might be heading that way.
  • lenghui - Friday, June 17, 2011 - link

    @L. and @semo,
    Thank you.
  • Movieman420 - Friday, June 17, 2011 - link

    Why on earth would SF/Ocz issue a recall when 99 percent of V3 owners have ZERO issues??? Atm they are looking for the proverbial needle in a haystack. Ocz support has reached out trying to find a customer with the bsod issue that is located near the SF bay area so they can send an engineer out to do a full workup for the first time on a known problem machine. Later the thread was removed as they most likely found someone. It's just a matter of time. All signs point to a power issue as the bsod occurs because the drive suddenly disappears because it loses power for whatever reason.
  • chrnochime - Friday, June 17, 2011 - link

    Your posts seem awfully supportive of them ....
  • sean.crees - Friday, June 17, 2011 - link

    I am so glad they have comments on these articles. I swear i learn more from people who comment on the articles than the articles themselves.

    OCZ almost had my money. Anand never seems to put much emphasis on instability issues with the non Intel SSD's. Sure he mentions it in passing in a sorta "ho hum, you don't really have to worry about this because the hand picked samples i get always work" sort of way. But when there are huge bar graphs you see him talk those up a lot about how that's all that matters.

    The problem is there should not be a stability vs performance decision. Stability is non negotiable. It should not even be part of the equation. It should be speed differences in different types of work loads and applications and thats it. You would never see a CPU review of AMD or Intel where they talked at the end about "well if your willing to give up some stability you can get more speed from this". NO WAY! They would say "this thing is not stable, don't buy it" because it doesn't matter how fast it is if it only works 90% of the time.
  • L. - Friday, June 17, 2011 - link

    99% of the OCZ drives have no issues it seems.
    There is a stability vs performance decision BECAUSE your OWN disk drives are NOT replicated locally and remotely.
  • michal1980 - Friday, June 17, 2011 - link

    really 99% have to issues?

    You have some data to back that up? Or just another OCZ troll.
  • DarkKnight_Y2K - Friday, June 17, 2011 - link

    You can go to OCZ’s Support Forum, count all of the unique Member Names that have signed up and posted a complaint about bsod issues with their drives and divide that number by the total amount of drives (Vertex 3s, Agility 3s, Solid 3s) sold.

    I believe the 1-2% is accurate but if OCZ is lying it’s probably not that difficult to find out the truth.
  • michal1980 - Friday, June 17, 2011 - link

    Which ignores all users that just returned the drive. Or used phone support, or used support on a different forum.

    again, that 99% no problem number is just made up.

    P.S. did ocz release the number of drives sold?
  • semo - Friday, June 17, 2011 - link

    OCZ are also running a very censored forum so some threads get deleted without an explanation. I also don't believe that arbitrary 1% figure.

    @OP. It took a very long time for Anand to even cover the 25nm transition fiasco. Even without a mass recall of all affected drives, Anand was satisfied with OCZ's handling of the issue.

    http://forums.anandtech.com/showthread.php?t=21433...
  • DarkKnight_Y2K - Friday, June 17, 2011 - link

    You have a valid point but would you just return drive or contact the manufacturer by phone? It’s possible but I believe most people in 2011 with SSD, HDD, memory, graphic card, motherboard, power supply, cpu problems are going directly to the manufacturer’s web site at least once to see if their issue can be resolved.

    I don’t know if OCZ have released worldwide sales figures but my general point was if they are lying someone with determination (competitors?) can find out the truth.
  • michal1980 - Friday, June 17, 2011 - link

    a good chunk of the competion is using the same chipset, they have little to gain by bashing OCZ.
  • e-drood - Friday, June 17, 2011 - link

    the sandforce 2000 series controllers have synchronicity & voltage transient issues beyond simple firmware update(s) resolution - the marchitecture has flaw(s)/omission(s).

    corsair acted quickly to recall to protect market rep AND establish "plausible denial of wrongdoing" should product defect litigation begin.

    ocz is AGGRESSIVELY deleting posts & threads containing any CRITIQUE of V3SF2000 ssd's & using HOSTILE online ocz staff forum comments to shift responsibility to any 3rd party chipset manufacturer AND to purchasers who are described/strongly implied to be technically incompetent and/or unable to follow simple installation instructions (ie: the multiple sets of 10-16 step instructions has reached the point of absurdity and INSULT).

    and the STAR PLAYER SANDFORCE has NOT issued single comment - first 1200 series controller disaster, NOW 2000 series controller mega-disaster -- THIS MARCHITECTURE DESIGN HOUSE HAS NO CREDIBILITY ANY LONGER! We Users Are Foolish To Support This Nonsense...
  • werewolf2000 - Saturday, June 18, 2011 - link

    I agree with Jeff72 ... I had similar problems with Vertex 3 240 GB (two of them). I also discovered too late that these liars from OCZ created a slower version of this SSD and I got such crippled ones.
    Never more anything from them.
  • Jeff72 - Saturday, June 18, 2011 - link

    OCZ Vertex 3 looks great on benchmarks. I wish I didn't have the blue screen issue on my Sandy Bridge system with them. I bought the OCZ Vertex 3 because of the great benchmarks. And after I had the blue screen issue on the first one, the benchmarks convinced me to exchange it at first in the hopes that it was an isolated incident instead of getting a refund right away. However, after getting the blue screen issue on a second new OCZ Vertex 3 SSD, and then seeing others having the same problem on the OCZ forums, and then seeing one of the OCZ forum moderators close a forum thread about the issue when people were trying to document the problem and troubleshoot and solve it, that's when I got the refund and went Intel 510 series 120GB SSD.

    This was my first SSD drive. I learned from the experience and value reliability and stability now much more than a bit more speed. Through this experience I now prefer Intel SSDs.

    As I mentioned before, if you want to get the OCZ, just get it from somewhere you can return it for a full refund if needed, in case you get the blue screen issue.

    I really hope OCZ can solve the issue and acknowledge it being solved and hopefully it is just a firmware fix. Their product works great when it works.
  • Movieman420 - Saturday, June 18, 2011 - link

    Yes my posts seem supportive...this is because I am a well informed member of the support forum that has taken the time to read past the first few posts of a thread and actually understand the why's and how's of the issues. @e-drood Staff deleting posts? Get real...at worse, some posts get moved to the complaints/recommendation section as they are purely complaints, not customers seeking actual support with an issue. One look thru the forum and you will find LOADS of complains about this issue...how can this be if Ocz is censoring their own support forum? Dude, you couldn't be more clueless if you tried. My advice..go and take the time to actually go thru and read, understand, and fully comprehend what is happening and why. And these 'disastrous and mega-disastrous' SF controllers out perform and out sell any and every comparable controller on the market..and every review you'll find, esp here at Anandtech, ranks them the best. btw...corair's recall is due to physically defective hardware with only the 120gb model. You really should get your facts straight before having a bash at things that you clearly do not understand.
  • Beenthere - Saturday, June 18, 2011 - link

    I don't want to get in the middle of a pointless pissfest but I can confirm that OCZ and other websites do remove some negatives posts and do ban some folks even when their posts are proper and appropriate. It seems like you really need to tip-toe around some website when you report a legitimate problem that should never exist if the product was properly qualified before being released to market.

    Certain hardware and software suppliers have a history of shipping half-baked crap and using consumers as un-paid beta testers. It's not difficult to see who these companies are as many of them are high profile and obtain glowing reviews even when the products are in fact documented to have defectives.
  • caseyse - Sunday, June 19, 2011 - link

    Well informed member of the support forum? I see 60 reviews for this drive at Newegg, and a third of these (17) provided a rating of 3 or less. Those that provided a rating of 3 were overly generous as I see they too are experiencing BSODs. These reviewer stats have not been censored and would appear to provide a better view of this drive's performance. I agree with others in this thread that reliability and safeguarding data is what's important, speed is a far distant second. This new controller was SF's opportunity to overcome its prior SF12xx product issues. This potentially could have a very serous outcome to both SF and OCZ.
  • e-drood - Sunday, June 19, 2011 - link

    "Tony The Tiger" OCZ UK Support has now posted Sandforce Forums "Heads Up - Check Your SATA3 Cables & Use Best Cables Re: BSOD's"!?!

    This is a joke and insult to buyers of OCZ V3 / V3 Max IOPs SF2000 SSD's AND OCZ doesn't give a toss about customer growing frustration with this newest and greatest Sandforce controller failure debacle AND Is It Too Much To Call This Fraud For Failing To Recall These SSD's From Market???

    The OCZ SSD's using Sandforce Series 2000 Controllers DO NOT HAVE STABILITY, ARE NOT RELIABILE & HAVE NON-FIRMWARE/MARCHITECTURE FLAW SATA3 INTERFACE ISSUES.

    Enough is Enough with this garbage hardware/SSD trash -- Issue Recall Now!
  • Coup27 - Monday, June 20, 2011 - link

    I've just spent the last half an hour reading various threads on OCZ's forum in their Indilinx, but mainly Sandforce section and the whole affair seems quite a shambles.

    Users are being asked to try all sorts of things which they should simply not be forced into attempting. I was reading the thread about them saying you have to buy quality cables to use on SATA-3 and it was deleted while I was even reading it.

    I do deffinately get the impression OCZ rely a lot on their user base to iron out problems for them. Intel drives may not have the benchmark scores on the graphs, but work they do. I have bought 3 Intel drives, one 80GB G2 for my personal machine and 2 40GB 320's for work and none have ever missed a beat in any way.

    As much as I love this website and Anand's articles personally, I am dissapointed by his response to OCZ's business practices and substandard validation procedure. The Spektek issue was not publically addressed until long after the horse had bolted and even then it was wrapped up in a lengthly article diluting down the message. And whilst articles like this exist which are one page suggesting there is a new midrange king in down, to not even update this article to warn readers that they maybe treated with a BSOD when they plug their AT recommended purchase into their new machine is very dissapointing.

    I have only ever purchased or recommended Intel drives and I cannot see that view changing at all.
  • xzynp7 - Tuesday, June 21, 2011 - link

    I will take the Intel drive. Intel firmware updates on the mail drive easily. Reliability is the most important criteria for me. The differences between these 2 drives are barely noticeable in real life use. I lost 500.00 on a similar high performer last year. Now all my drives are Intel SSD's all reliable and fast enough.

Log in

Don't have an account? Sign up now