A closer look at fresh versus used SSD performance

Many would argue solid-state drives are the future of PC storage, but some of today’s SSDs apparently aren’t as resilient as one might expect. Drives based on Indilinx’s “Barefoot” IDX110M00-LC controller are a case in point. In a recent SSD round-up, we discovered that although Barefoot-based OCZ Vertex and Super Talent UltraDrive ME drives excelled in synthetic tests like HD Tach and IOMeter, they suffered from surprisingly poor real-world write and file copy performance. Our results were consistent and repeatable within the confines of our test environment, so we probed deeper in an attempt determine the true nature of the issue.

We initially tested these drives with revision 1275 of Indilinx’s firmware—their shipping configurations. Curious to see whether the latest firmware would affect performance, we secure-erased the Vertex to put it into factory-fresh form, flashed it with the most recent 1370 firmware, and ran it through our test suite once more. The newer firmware didn’t affect the drive’s performance in any of our tests, though.

Indilinx’s Barefoot SSD controller

Up to this point, we had tested all of the SSDs in a used state. We secure-erased each drive before its run through the gauntlet, but the first test in our benchmark suite—HD Tach’s full disk benchmark, which we run three times—fills all flash pages by writing across the entire drive.

Testing SSDs in a used condition is important because of the block rewrite penalty associated with flash memory. In an SSD, flash memory is split between 4KB pages organized into 128-page, 512KB blocks. Empty pages can be written to directly. However, occupied pages can only be written on a block-by-block basis, which involves reading the contents of the block, modifying the necessary pages, and rewriting the block as a whole. These extra steps take time, resulting in a performance penalty.

Deleting files simply marks pages as available, rather than actually clearing their contents, so any SSD will encounter the block rewrite penalty after prolonged usage. Even a drive that, cosmetically, shows plenty of available storage capacity can be out of empty pages. Once in this used state, every write request will be slowed by the block rewrite penalty.

We believe that testing SSDs in this used state is the best way to evaluate their performance. A solid-state drive may post impressive benchmark scores when it’s fresh from the factory, but it’s more important for SSD controllers to deal gracefully with a lack of empty pages. As we discovered, some do a much better job than others.

We briefly explored the performance of used versus factory-fresh SSDs in our initial round-up, comparing the 4k random write performance of each drive in each state. The Indilinx drives showed no obvious signs of weakness; response times slowed when the drives were used, but by a lesser degree than drives powered by Intel or Samsung controllers. In fact, in a fresh state, the Barefoot drives had the quickest response times of the lot.

To determine whether their used state was responsible for the Indilinx drives’ poor file copy performance, we ran a handful of freshly-erased drives through iPEAK and FC-Test. These drives were tested with the same system configuration as our initial round-up, this time with the OCZ Vertex updated to Indilinx’s 1370 firmware revision.

Our iPEAK results represent an average service time across all nine of our custom workloads, four of which feature file copy operations. In a factory-fresh state, the Indilinx-powered Vertex isn’t substantially slower than the Intel drive or the Samsung-powered Corsair. However, the Vertex’s service times are nearly four times longer when the drive is in a used state. The Intel and Corsair drives slow down, too, but only to about twice their original service times.

Next we’ll look at some real-world file creation, read, and copy operations courtesy of FC-Test. Here we’ve presented the average throughput across all five of the app’s test patterns, which contain a mix of large and small files, and large and small numbers of files.

The Vertex offers quicker file creation speeds than the X25-M when both are in a fresh state. However, while the Intel drive’s performance only drops by about 20% in a used state, the OCZ’s falls by nearly 70%. The Samsung controller inside the Corsair P256 doesn’t particularly like being in a used state, either; its file creation performance drops by nearly 50% going from fresh to used.

In theory, flash memory’s block rewrite penalty should have no impact on read performance—used drives should read as quickly as fresh ones. And they do, at least with the X25-M and P256. However, the Vertex’s read performance when used lags behind that of a fresh drive by about 10%. That’s not a huge margin, but it’s a performance gap that, as far as we know, shouldn’t exist at all.

Copy tests combine read and write operations, and as one might expect, the Vertex doesn’t fare so well here. Its used performance is about 60% slower than a fresh drive, which is a much bigger margin than either the Corsair or Intel drives. The former’s performance drops by 30%, while the latter is only 15% slower.

The Indilinx controller would appear to have performance problems when dealing with drives in a used state. But is the controller to blame? Our testing was conducted with Windows XP Service Pack 2 and an older 5.1.0.1022 AHCI driver revision for our system’s ICH7R south bridge chip. To make sure we weren’t missing any critical software updates for our test platform, we brought the OS up to Service Pack 3 with all the latest hotfixes and installed the most recent Intel 8.8.0.1009 AHCI drivers, in addition to new chipset drivers. Once more, we secure-erased the Vertex and ran it through our benchmark suite, only to find that this new configuration’s performance was essentially identical to that of the old one. Even with the latest drivers, firmware, and XP updates, the Indilinx controller’s used-state performance in file creation, read, and copy operations was still rather poor.

All of this testing was conducted on a relatively old system based around a Pentium 4 processor, 955X chipset, and ICH7R storage controller. While that’s hardly cutting-edge hardware, this very same test system had no problems wringing excellent performance from SSDs based on Intel and Samsung controllers.

Could our use of Windows XP explain the Indilinx controller’s woes? Perhaps. We’re currently conducting Vista testing to see how these SSDs fare. Other sites have also reported better FC-Test performance from the Vertex running under Vista x64. However, based on some initial results of our own Vista testing, we remain skeptical. And the problem with XP performance remains.

Exploring the problem
The block rewrite penalty affects all flash-based SSDs, but as we’ve discovered, some do a better job of managing this problem than others. Indilinx’s Barefoot controller appears to be much worse off on this front, at least when running on Windows XP, than controllers from Intel and Samsung. But why?

One potential problem is Windows XP’s default 63-sector offset, which starts the first partition right in the middle of an SSD page rather than at the beginning of one. Since flash can only be written in 4KB pages, the overlap caused by an initial misalignment can result in more pages being written than necessary for a given write request. The performance implications of this situation are particularly troubling when the drive has no empty pages. Microsoft suggests misalignment can drop SSD performance by up to 50% when block rewrite penalties come into play.

One solution to XP’s default partition offset is to create partitions manually with a custom offset. Doing so is possible using a Diskpart utility available from Microsoft. Such tweaking isn’t necessary for all SSDs, though. When asked whether the X25-M takes XP’s 63-sector offset into account, Intel told us that its SSD tech is alignment-agnostic, just as long as there’s some alignment present in the operating system. Also, the Samsung controller doesn’t seem to be as adversely affected by XP’s default offset as the Indilinx chip in the Vertex.

In fact, custom partition alignment may not even be necessary with the Vertex. When asked about the issue, OCZ Vice President of Technology Development Michael Schuette also suggested that an SSD could side-step XP’s 63-sector offset by employing a default 128KB offset in its firmware. So why hasn’t such an offset been incorporated into the Vertex’s firmware? Because that firmware comes from Indilinx; according to Schuette, OCZ doesn’t have access to the source code.

We fired off an email to Indilinx regarding the issue, specifically asking about XP’s 63-sector offset and the comparatively poor used-state performance we observed from the Barefoot controller:

I’ve been testing a couple of Indilinx-powered SSDs (OCZ’s Vertex and Super Talent’s UltraDrive ME) and have discovered that the drives offer much lower performance in a used state than when in factory-fresh form. This is of course an issue for all SSDs, but the Indilinx drives suffer a much greater performance drop from fresh to used than SSDs powered by Samsung and Intel controllers.

I’ve been testing with Windows XP, whose 63-sector offset is apparently to blame for the Indilinx controller’s disproportionally poor performance in a used state. Does the Indilinx controller take XP’s 63-sector offset into account? Is it possible to add a default offset to the controller’s firmware so that XP users don’t have to resort to manual partition alignment for optimal performance?

Indilinx’s replied with the following:

We support TRIM Command to recover the performance drop after fragmentation(used).

This command shows the original performance like Intel.

Rather than directly addressing XP’s 63-sector offset and the controller’s poor used-state performance, Indilinx is essentially suggesting that we avoid a used state altogether. The TRIM utility erases any pages marked as available but occupied with old data, reducing the chance of a block-rewrite penalty to slow a write request. This program effectively pushes a drive toward its factory-fresh state without erasing the data you actually want to keep.

That all sounds well and good, but Indilinx’s TRIM utility doesn’t appear to be ready for prime time. In a post releasing a beta version of the app in its SSD forums, an OCZ rep notes that he repeatedly encountered data corruption when running the utility on a 64-bit Vista system with an ICH10R south bridge. In fact, although the app is called 100% safe for 32-bit operating systems, OCZ says it’s only “looking around 50% safe” for 64-bit operating systems. Support for 64-bit Vista and Windows 7 is listed as “limited”, and Windows XP 64-bit isn’t supported at all. Incidentally, we tried the latest TRIM utility on our original test system, a 32-bit XP machine, and it didn’t work. OCZ’s suggestion was to run the app up to six times, rebooting after each.

Windows 7 promises native support for the TRIM command, but the OS hasn’t been finalized, and Indilinx’s home-brewed utility doesn’t look like a reliable solution for those running other operating systems. Based on the accounts of OCZ’s own SSD forum admin, I wouldn’t trust the app with a 64-bit operating system. The fact that data corruption has been observed would make me wary of running the utility on a 32-bit OS, as well.

Although the TRIM utility clearly isn’t a robust solution, it appears to be the only one Indilinx is offering to address the poor used-state performance of Barefoot-based drives in Windows XP. And before you argue that Windows XP is far too antiquated to matter, keep in mind that it’s still widely used by PC enthusiasts and mainstream folks alike. If Intel and Samsung have more effectively addressed used-state performance with their SSD controllers, shouldn’t we expect the same from Indilinx or anyone else hawking an SSD controller?

We asked OCZ’s Michael Schuette how much pull the company has with Indilinx in terms of getting new features (like a default partition offset) introduced to firmware and were told that OCZ is “doing a huge part of the debugging and performance analyses” related to the drive. Schuette went on to say that, “With more access to the details on the hardware and firmware, I am sure we could make this controller really fly, it is pretty good already, but we are stuck behind the iron curtain.”

Indilinx’s Barefoot controller appears to be, charitably, a work in progress—as are the drives based on it. At present, their used-state performance issues have not been addressed adequately. Indilinx may be the only party capable of rectifying the problem via firmware changes, but that doesn’t absolve OCZ, Super Talent, and others selling Barefoot drives of responsibility for a glaring flaw in the products they are selling to consumers.

If you want to dabble with Windows 7 release candidates, fine-tune partition alignments, or roll the dice with a TRIM utility that could corrupt your data, SSDs based on Indilinx’s Barefoot controller certainly have intriguing potential. However, for those seeking a solid-state drive that doesn’t require extensive tweaking for optimal performance, products based on the latest controllers from Intel and Samsung are much safer bets—including, potentially, OCZ’s own recently introduced Summit series.

In highlighting the used-state performance penalty associated with the Barefoot controller, we’ve focused exclusively on Windows XP performance with our ICH7R-based storage test system. But that’s not all we’ve been doing. A new round of SSD testing has already begun on an updated system built around a Core 2 processor, ICH10R south bridge, and Vista x64. Stay tuned for more.

Comments closed
    • Dirge
    • 12 years ago

    Could the next test include Linux file systems such as ext3?

    • Wickedt
    • 12 years ago

    @ALL

    I am not sure why, but their is something drastically wrong with my OCZ Vertex 120 Gig drive. After using it normally for the last 4 months, running the wiper tool they supply on it once a week, these are the results i am getting from a used system, and 64 bit vista to boot!

    §[<http://s665.photobucket.com/albums/vv20/oczuser/?action=view&current=results.jpg<]§ This is a used drive, so why are my results so impressive? Like i have said many times, this is what most (~90%) Vertex useds are getting. The results of this site are just plain wrong. and are flawed. I do not like doing these benchmarks, because they are hard on the drive, because it is MLC nands, and any research will show you this. The drives are meant to be enjoyed, not benchmarked 24/7. This people is a real world test, on a used system. I think a lot of you may want to read this also, it is very relavant to all this, contrary to what the site says. §[<http://blogs.msdn.com/e7/archive/2009/05/05/support-and-q-a-for-solid-state-drives-and.aspx<]§

      • derFunkenstein
      • 12 years ago

      If you’re running the wipe util, your drive is not in a used state anymore.

      • Hell Racer
      • 12 years ago

      Trim tool (wiper) is *[

      • pmonti80
      • 12 years ago

      Ok, I don’t like the conclusions of both reviews either (waiting for the 3rd article). But this post is just …….., OCZ Vertex has been out for 3 months or so, not much more, and the wiper tool has been out for less time. So the whole post is probably done by someone with a little bit too much imagination.

        • Wickedt
        • 12 years ago

        Ok so possibly 4 months, bottom line is, look at the results instead of being a smart ass, if what they said in this review is true, then i should not have those results. Your nitpicking, WTF does imagination have to do with a screenshot? This is proof, in case you do not understand, thats why screen shots are considered valid proof.

        Wheter ya want to beleive it or not, or think it is made up, or whatever you think, i really could not give a damm. This people is a 120 Gig OCZ vertex.

        Do yourself a favor, go to the OCZ forum and look at the benchmarks people are sending in, mine are actually lower than most.

        You people in this forum floor me with your comments, people wake up, it was a crappy review!

          • Hell Racer
          • 12 years ago

          Most of people post screenshots in unused state, or they use that beta tool. My bench results are nice too. But that doesn’t lead to the fact the review is crappy, it just made from another point of view, where wiper still doesn’t exist yet (because it’s beta), and there’s no Windows 7 (because it isn’t released yet).

            • Wickedt
            • 12 years ago

            Actually Win7 is RC…release candidate, and has trim native to the os, win7 also turns off indexing and prefetch when it detects a non mechanical drive, or SSD.

            We are just waiting on Indy to realease the new FW for native trim support.

            • Convert
            • 12 years ago

            Everyone always seems to confuse RTM and RC.

            Win7 is not out yet. It’s final version is not out yet. It’s still buggy. It’s not a finished product yet. How many times does a person need to say this.

            • tfp
            • 12 years ago

            Every time the topic is brought up until Windows 7 is released.

      • jackaroon
      • 12 years ago

      I’m really happy for you, but I’d have to upgrade my OS and most of the components in my PC to get those results. I just want a fast drive.

      Yes, I AM just jealous.

    • nerdrage
    • 12 years ago

    SSDs remind me a lot of LCDs: lots of benefits compared to the previous gen technology, but at the same time a lot of new drawbacks/weaknesses that the previous tech did NOT have.

    • Convert
    • 12 years ago

    I will only say that I own an x25m and I have seen an improvement on boot times. Those are the only charts that I noticed were a bit odd, otherwise everything else seems to jive with what I can reproduce.

    • Trymor
    • 12 years ago

    Shouldn’t hardware manufacturers know MORE than tech review sites? It seems more and more often that isn’t the case…

    EVERY manufacturer better start PAYING ATTENTION rather than blowing off these review sites such as our beloved TR. If not, blame no one other than yourselves when you fade into obscurity…

    Edit: At least in the case of Indilinx, OZC at least seems to be listening.

    • Chloiber
    • 12 years ago

    New TRIM-Tool has been released by the way. So far no data corruption (also on x64 😉 )

      • Chrispy_
      • 12 years ago

      Cool, must be ready then 🙂

      I wish all manufacturers would use sample sizes of one like you just did. It sure would alleviate all those pesky development delays caused by unnecessary testing.

      • Freon
      • 12 years ago

      So how much is OCZ paying an hour for all these beta testers out there? At least it is a “real world” test!

        • Chloiber
        • 12 years ago

        As much as Intel is paying or any other hardwareseller releasing beta-software – nobody is forcing you to use it.

        I don’t quite get it what your problem is (not talking specifically about you). Yes, there is a problem with performancedegradation. Maybe a bigger one than on other SSDs. But EVERY MLC-SSD got that problem and every MLC-SSD is losing performance over time. What we are seeing from Indilinx (not OCZ – everything comes from Indilinx) is an offer for everyone who can’t wait for working W7 Trim. You don’t have to use it if you don’t want to. Other SSD manufacturers aren’t offering anything and if you want your performance back with such an SSD you will have to SecureErase the whole thing.

        I am aware that the Indilinx drives are losing more performance than others. But WE got a solution and the other’s have none. As I said. Everyone using x32 system is at 100% performance 100% of the time, 100% secured. And nearly everyone using a x64 system too. If you think it’s too risky, don’t use it, end of story.

          • Freon
          • 12 years ago

          Intel seems to be pretty far ahead of the curve. I guess the only question remains, if you buy an X25-M right now will you get the newer firmware that came out a while back? If so, I’d say it is pretty compelling as a no fuss drop-in solution. Sure it degrades with use, but much more gracefully in more configurations.

          Unfortunately good engineering isn’t cheap. For the premium though, I’d rather pick up an X25-M over a Vertex and not have to worry about a 40 step setup process, having the exact right OS, etc.

          I’m not buying any SSD right now. However current performance of the X25-M looks compelling across the widest range of configurations with the least amount of special setup, and I would buy one if the price were better. I don’t think I could say the same about any other SSD right now.

            • indeego
            • 12 years ago

            According to newegg reviews people are still getting new intel drives with the old firmwareg{<.<}g Will take a while likely for intel to change their mechanism for updating drive firmware, validate, and push into channel. I'm glad I waited on SSD. Still looks like it has a bit of early adopter syndrome going on. I will build around SSD and HD for my Feb 2010 buildg{<.<}g

            • bats
            • 12 years ago

            g{

            • Ushio01
            • 12 years ago

            “the firmware update is a bootable cd iso imag

            Where is a location for a step by step guide for the above as i have no idea how to set one up.

            • Convert
            • 12 years ago

            §[<http://downloadcenter.intel.com/Detail_Desc.aspx?agr=Y&DwnldID=17485<]§ Let us know if the included instructions aren't clear. My brand new Intel drive from newegg came with old firmware just to share.

    • GreatGooglyMoogly
    • 12 years ago

    I was with you until the abrupt leap to the partition alignment theory, without any actual test to back it up. I guess the article was rushed, which is understandable.

    • mi1stormilst
    • 12 years ago

    Looking forward to even more helpful information in the new article.

    • Krogoth
    • 12 years ago

    SSDs have never been the same beast as magnetic-based HDDs.

    Performance and capacity degradation happens to be one primary drawbacks of SSDs. The current issue is ironing out the rate of degradation to more an acceptable level. It does not help that most customer-grade OS’s IO kernel are still coded with HDDs in mind.

    I see no point in getting SSDs until the issues are ironed out and GB/$$$$ ratio comes closer to HDDs.

      • bats
      • 12 years ago

      g{

        • Krogoth
        • 12 years ago

        SSDs have a few performance degeneration problems that are unique to themselves. Flash bit decay. SSDs has to store “actual” data in fixed “blocks”. TR article investigates how well current drives handle this problem).

    • ronch
    • 12 years ago

    With platter-based hard disc drives being cheaper than ever these days, I really don’t think it’s very sensible to get an SSD now unless you have a gold mine. I personally would wait a few more years. By then SSDs could probably be on price parity with platter-based HDDs and have the issues worked out.

      • mesyn191
      • 12 years ago

      Current mech. hard drives are indeed perfectly fine, these SSD’s are supposed to be, and with a little tweaking can be, quite a bit faster though. Mo’ performance = mo’ bettah.

      Personally I’m holding off until they get cheaper and the bugs get worked out. If one of the news articles from yesterday’s roundup is to believed than we should only have to wait a few more months or so for that to happen.

    • mattthemuppet
    • 12 years ago

    all I’m getting out of these reviews is that if you want a consumer level SSD, get an Intel X25-M 🙂

    • boing
    • 12 years ago

    I don’t get why the Vertex is so popular amongst many enthusiasts. Is it particularly cheap? I’ve never seen it excel in any benchmarks compared to its competitors.

      • MadManOriginal
      • 12 years ago

      I think it is because it’s relatively cheap in GB/$ compared to the only previous ‘this is the SSD to get’ Intel X25-M (forget about X25-E) and got a bunch of glowing reviews plus hype around the Indilix controller after JMicron controller SSDs were shown to be crappy.

      • Chloiber
      • 12 years ago

      Because “enthusiasts” mostly don’t have any problems with them.
      I agree though that the Vertex is overpriced compared to the UltraDrive (128GB is about 2.5 vs 3.5€/GB!).

      Look at this poll (german forum) :
      §[<http://translate.google.com/translate?prev=hp&hl=de&js=n&u=http%3A%2F%2Fwww.hardwareluxx.de%2Fcommunity%2Fshowthread.php%3Ft%3D613183&sl=de&tl=en&history_state0=<]§ 54 out of 58 ppl ar satisfied with their Indilinx-based SSD. And: §[<http://www.laptopmag.com/review/storage/samsung-ssd-256GB.aspx?page=2<]§ Indilinxdrives aren't as bad as you might think after reading this article. I don't think the article is bad or anything, but it's just 1 of many out there - so don't base all your informations on this one (neither on any other). I agree that SSDs aren't quite ready for the masses, but "enthusiasts" or ppl who "understand" their computers know about the problems, heard about the solutions and have no problems at all.

        • MadManOriginal
        • 12 years ago

        Your last sentence is kind of funny because in some cases even the “manufacturers” (well, really they’re almost like rebadgers although some do stuff on their own) don’t seem to “understand” the product!

        • derFunkenstein
        • 12 years ago

        Your use of quotes with “understand” is probably the most truthful thing you’ve written. People who think they know what they’re doing buy SSD tech before it’s ready for prime time.

        • Freon
        • 12 years ago

        How many people even realize they’re not getting the best performance? You could probably sell 5400RPM drives as 7200RPM and many people would never notice the difference.

        Polls asking, “are you satisfied?” aren’t very useful because that’s really a subjective question. And since initial performance is still so good, people may not realize the issue so they buy a drive, use it for a few weeks and it is great, and answer “OMFGBBQAWESOME” on random polls on the internet.

          • Chloiber
          • 12 years ago

          I can guarantee you that every single one using that forum is aware of the problems with Indilinx SSDs.

          Your information is based on this one review. If you read it, you would think after you plug in your ssds you will see about 100 BSODs and lower performance than on an HDD. This is not the case. You never used the drive, you can’t tell if it’s working well or not.
          As I said before: it’s good that you exposed problems with the SSDs in this review, but I’m still thinking that you are exaggerating. This isn’t fanboyism, it’s the way it is, if you are looking at other reviews and comments from users who actually bought such an SSD and are using it everyday.
          That’s a poll from users who are 100% aware of the problems, it’s an open forum, no manufacturer board and yet you want to tell me that the drive sucks and nobody should buy it because you will only have problems with it.
          And I’m telling you that nearly every buyer is satisfied with the product – what else do you want? That I’m going to do an interview with every person posted in that thread?

            • Freon
            • 12 years ago

            You didn’t address that the poll does not expose any actual objective, scientific, reproducible data. It’s a poll on a very general subjective question. It’s not remotely a good basis for an argument. We’re not discussing a subjective topic, like art or music.

            TR’s tests do reveal objective data, that in theory anyone with the proper equipment could reproduce. Sure it is just one review, valid point. It’s good to look at many reviews and try to figure out where issues lie based upon the differences in test setups. I’m not going to discourage anyone from looking at other independent reviews, though I do warn that I don’t feel all reviews have equal quality. I generally trust Anandtech, for instance.

            Whether or not the people answering the poll know about the Indilinx issues is ancillary to the core issue with using such a poll as the basis for a position or argument. I reiterate, it is not!

            I’m not really concerned if some random people on the internet are satisfied with it. People are often satisfied with placebo medicine, too. It doesn’t mean there is actual efficacy of the product. This really doesn’t prove a darn thing. This is just not an acceptable debate tactic, so don’t expect to sway me or many others here with this sort of rhetoric.

            “If you read it, you would think after you plug in your ssds you will see about 100 BSODs and lower performance than on an HDD.”
            “…you want to tell me that the drive sucks and nobody should buy it because you will only have problems with it.”
            etc…

            The strawman/hyperbole routine is also not a valid debate tactic. Perhaps we should start back at square one with a review of logical fallacies? I don’t think we’re making much progress here.

            • Chloiber
            • 12 years ago

            The only thing I wanted to tell you is that nearly no user got a problem with his SSD. I am not referring to the performance degradation-part of this article (basically page 1), I am referring to page 2, which isn’t objective at all.

            Edit:

            But yes, let’s leave it. I agree with you that a poll in a “random” board (for you) can’t be scientific. It’s hard to show you how the ppl using that board actually feel about the indilinx-drive, because obviously you don’t know them/don’t speak their language – was a bad idea to link it, I agree 😉

        • indeego
        • 12 years ago

        /[<"Look at this poll (german forum) : 54 out of 58 ppl ar satisfied with their Indilinx-based SSD."<]/ This isn't scientific. In fact, it is anti-scientific. §[<http://en.wikipedia.org/wiki/Opinion_poll#Potential_for_inaccuracy<]§ Don't spread crap like this around and pass it as anything near accurateg{<.<}g

          • Chloiber
          • 12 years ago

          Sorry, I missed the part where 1 review on 1 setup with 1 drive is scientific.

          And I also missed the part where I said it was scientific and the absolute truth.
          If I would have wanted to imply that that poll is scientific I would have said “93% are satisfied”. I explicitly said how many ppl of how many in total are on that specific site. Maybe you should stop quoting wikipedia and learn what the true reason is for a poll to be unscientific.
          So please stop talking about that fucking poll.

          I even posted the link, I didn’t even comment it. You started the whole discussion about that poll O_o
          If you don’t want to talk about it, then just stop talking, I didn’t want to :>

            • CampinCarl
            • 12 years ago

            Actually, it is. It’s called ‘bug testing’–you don’t just do the every day cases, you go for random extremes and see if something breaks. It broke. Bug found.

            • Freon
            • 12 years ago

            “Sorry, I missed the part where 1 review on 1 setup with 1 drive is scientific.”

            I think you have some serious misunderstandings about what is and isn’t science.

            Any one test on one setup with one drive can be scientific. If performed correctly it should be reproducible by anyone with the same setup, same drive, and using the same procedure. Sample size is irrelevant to determine whether it was scientific or not. Good science in fact does narrow tests down to specific circumstances. You don’t want to convolute your test.

            It may be irresponsible to claim in a conclusion that the test is representative of what everyone will experience with the drive, but that’s a separate issue and doesn’t mean the test or data were scientific or not. It means the conclusion was irresponsibly written. This would be the broad brush or hasty generalization fallacy. It would go too far to say everyone will experience the same thing as the test shows.

            • Chloiber
            • 12 years ago

            I agree. I got some difficulties writing what I want to say as english isn’t my main language.

            I think I forgot adding “considering” before “review”. And that’s what many ppl are doing here. Or replace scientific with representative.
            But well whatever – I think I’m done here 😉

      • Trymor
      • 12 years ago

      Because you can get an OCZ branded $100 SSD with much better performance (at least as a boot/system drive) than a (edit: older, ready to be replaced) mechanical drive, or another ‘cheap’ SSD.

      At least I have seen that happening in some cases, and thats what I think about the situation currently 😉

      • mutantmagnet
      • 12 years ago

      Vertex is popular because of Anandtech and Jmicron.

      You see when SSDs came out there was only Intel, Samsung and everyone else based on Jmicron controllers. Jmicron controllers were so bad that in a used state SSDs could perform worse than platter hard drives.

      Anandtech at some point published the most indepth dissertation on why this was happening. Before they published it they shared their findings with other SSD manufacturers who didn’t seriously heed their advice and warnings. Initially neither did OCZ but that was because they were viewing performance differently.

      Once they realized Anandtech and themselves weren’t going to see eye to eye on the issue they decided to give Anandtech’s ideas a shot and made the Vertex based on Anadtech’s suggestions and the Apex based on the old school thinking of how SSDs should work.

      As a result Anandtech inserted glowing (and I do mean bright enough to make angels jealous) praise on OCZ in their paper and backed up their praise with results that showed Vertex being in a class of its own (though it wasn’t even close to being as good as an Intel).

      As a result OCZ has dominated the hearts and minds of enthusiasts for SSDs because they were the first company to offer a true alternative to Intel’s higher priced offering and showed they cared about doing the right thing for their consumers.

      The Vertex only looks bad here because its Indilix controller is being compared to the Intel controller (which was always good) and Samsung (the only SSD manufactuer back in the day who showed potential of matching the expertise of Intel but didn’t do so until recent with the P256)

      The Vertex isn’t bad at all. It’s just the baseline for where all SSDs should be considered good enough to be purchased.
      Anything below that is garbage which is the majority of consumer class SSDs. (Don’t know about enterprise class)

        • GourdFreeMan
        • 12 years ago

        mutantmagnet, your comments about Anandtech aren’t, strictly speaking, accurate. You seem to be conflating Anand’s unabashed praise of Intel’s SSDs with the Vertex in your recasting of history.

        Anand was the first professional reviewer to comment on the JMicron controller’s poor performance (though admittedly a blog writer caught the issue months before Anand). He did so in a review where he praised the performance of Intel’s then newly released SSDs and offered them glowing praise. When the Vertex was released he basically said it was acceptable though not nearly as good as Intel’s drive. To be fair, he did toot his own horn about getting OCZ to use his feedback in tuning the Vertex firmware, when there was already a weeks old poll on the OCZ forums demanding the same thing. He has also done his best to ignore Samsung SSDs (using old models in his reviews, and never commenting on their performance vis-a-vis random write performance) despite the fact other reviewers (like here at the Tech Report) have praised them.

          • mutantmagnet
          • 12 years ago

          This is just a matter of interpretation.

          My recasting is biased because the moment I saw that section on OCZ I knew it was the kind of positive endorsement that would catapult them to the front. Nothing has happened since then has refuted my assumption.

          I can see why you would feel they weren’t offering themselves as much praise to OCZ compared to INtel or their skills as analyzers and you are technically correct.

          But there’s always more to any form of communication in what you say. How you say it matters.

          The way Jmicron was correctly eviscerated and Samsung was unfortunately neglected could only put OCZ in the best light possible because Intel is just too expensive for a consumer class product.

          I thank you for adding some additional insight into the matter because I wasn’t aware of the possibility that the OCZ community was requesting OCZ directly for the type of changes Anandtech pointed out. It helps create a more complete picture of what lead to this shift in the SSD market back then.

            • wibeasley
            • 12 years ago

            “l[

    • BmUs
    • 12 years ago

    A Simple request to TR:

    I think a missing link here is the boot time. This new article doesn’t explain at all why boot times are slower with the Intel SSD’s then the 7200 mechanical HDD.

    If we read the charts correctly, even if we take that the boot time is a 100% wright and no read, on the Intel SSDs, the boot time should not be impacted as much as it did by the “used” state.

    Using the chart for both read and write operation test (# of writes being equal to #reads) we see a drop in speed of only 15% on the intel SSDs. That means that in the original benchmark, the SSD in a fresh state would have booted in ~38s, only 4 sec faster then a 7200HDD…something is not adding up.

    Could you please measure boot times in the same manner as the original article, but in a fresh vs used state?

    Thanks.

      • Waco
      • 12 years ago

      Agreed. The boot times were the only real problem I had with the original article and this update doesn’t address anything except the whiny fanboys. :/

    • Ushio01
    • 12 years ago

    Having read articles here and on AnandTech about SSD performance degradation over time, what kind of usage scenario is actually necessary to run into these problems as I’ve been using an SSD for about 9 months as an OS and AP drive with no slowdown or performance issues.

    The only thing I have noticed is the extraordinary performance increase to everyday tasks which even significant processor upgrades have never given my system in any noticeable way comparable to that given by my upgrading to an SSD.

      • Chloiber
      • 12 years ago

      What SSD are you using? Depends (as you obviously can see in the article).

      I had big performance drops when I’ve written about 1.5x to 2.0x times the size of the SSD.

      By the way:

      §[<http://www.ocztechnologyforum.com/forum/showthread.php?t=57254<]§ Statement from Tony (Hi Tony! :) I will quote an important part regarding trim: l[

        • wibeasley
        • 12 years ago

        How do you estimate how much you’ve written? Is there a utility that tracks that?

          • Chloiber
          • 12 years ago

          I don’t know of any…it’s really hard to tell in “all day use”. I just tested it with FC-Test.
          I just look at CrystalDiskMark every week and then I’ll see when I need to use trim.
          But at the moment trim runs automatically every 4 days. Don’t need to worry anymore 😉

        • derFunkenstein
        • 12 years ago

        Tony also refers to Microsoft as “M$” which takes either big balls or tiny brains when you work for a hardware vendor. Kinda kills his credibility.

        Unless there is integrated TRIM support added (backported) to XP and Vista, it’s a permanent fix for 100% of the computers out there now and will remain so for everyone who doesn’t immediately jump to Win7.

    • deepthought86
    • 12 years ago

    Windows XP has a 62% market share. Vista has a 23% share. Based on this and the results of this review it looks like SSDs are nowhere near ready for mainstream use.

    The bleeding-edge crowd doesn’t seem to realize this. Some idiots want to move over to Windows 7 testing even!

      • MadManOriginal
      • 12 years ago

      Well yeah you’ve got a point but what’s important is the market share among likely buyers of retail add-in SSDs. I am sure that a lot of XP boxes are corporate boxes and those are unlikely to be considered for SSD upgrades.

    • DancesWithLysol
    • 12 years ago

    In a related topic to all this controversy with current SSDs: perhaps it is time for a new file system. NTFS (and almost every other file system in use today) has been optimized for rotational media and its performance characteristics. I imagine if file system engineers were to example flash memory’s performance characteristics and build a new file system intended for SSDs, you would get a very different file system implementation.

    I’m sure people are working on this right now, in fact.

    • crazybus
    • 12 years ago

    Great article. It’s apparent that consumer-level SSDs are still very much bleeding-edge tech and there are still kinks to be ironed out. Anytime you see firmware revisions drastically changing performance, such as the Intel X25-M’s used performance or Anand’s Indilinx drives, you know you’re dealing with products not quite ready for the mainstream. Hopefully by this time next year these types of issues will be rare.

    Still, I think that if, for whatever reason, you insist on using an SSD with an older operating system such as XP, it’s really in your best interest to investigate the tweaks necessary for optimal performance. I understand the desirability of “plug and play” hardware, but keep in mind XP came out nearly eight years ago and is tuned for mechanical disk drives, sometimes to the detriment of SSDs. There still seems to be some dissension on how much abstraction there should be between the OS and the disk controller. “Smart” SSDs such as the X25-M can maintain good performance despite less than optimal conditions, but bring a degree of “black box” mystery. This separates the men from the boys in the controller department, but I can’t help thinking it’s not an ideal scenario.

      • deepthought86
      • 12 years ago

      /[< if, for whatever reason, you insist on using an SSD with an older operating system such as XP<]/ XP has 62% market share, Vista has 23%. Regardless of age your argument makes no sense. If anything, ONLY XP should be tested.

        • crazybus
        • 12 years ago

        Of course testing on XP is still relevant for a section of the market. I just recently finished setting up XP on an Eee PC 901’s SSD. Your market share numbers fail to mention that the majority of PCs are never upgraded. The small enthusiast market that may install a SSD in an older system is minuscule in comparison to the OEM/business/Buy More market.

          • jackaroon
          • 12 years ago

          The point about PC’s that never get upgraded is historically true, but SSDs could be “special” in that context. If SSDs were to hit that magic goal where they are “just like a regular hard disk, except fast” (and cheap enough) it *would* be the one upgrade that is probably worth it to anyone, not just gamers, number crunchers, etc, who are used to upgrading all their components regularly. Everyone waits for the hard disk, sometimes.

    • DancesWithLysol
    • 12 years ago

    I think TR could put a lot of these OCZ-forum-users criticisms to rest by simply installing Windows 7 RC and measure what the fresh vs used delta is there. If the Indilinx controller continues to perform poorly with Win7’s TRIM support Indilinx has zero arguments left.

    Plus I’m personally curious what the performance is like under Windows 7 because that is what I’ve been running since the Win7 beta was released on Technet. For me, Vista and XP are just historical curiosities. 🙂

      • wibeasley
      • 12 years ago

      I bet you’re right that those criticisms would be addressed with Windows 7. But I’m not convinced it’s TR’s responsibility to focus solely on their criticisms. That’s not to say I’m uninterested in reading Windows 7 results.

      If they used Windows 7 and unexpected results were still seen (with any of the drives, not necessarily OCZ), do you think we’d dismiss it because a pre-release OS was used (especially boot times)? From the podcast, it sounds like they’ve intended to switch to a Windows 7 platform for testing drives once the OS is released.

    • wibeasley
    • 12 years ago

    It’s ironic that the intention of some of the OCZ forum members (like jwilliams and techspec6) could backfire. From what I can tell from their foum (http://www.ocztechnologyforum.com/forum/showthread.php?t=56829), they wanted unhappy OCZ members to come to TR and criticize and bully Geoff. But that plan could blow up in their hands if some of them appreciate the perspective of a neutral author that’s not tied to any vendor.

    I think their own insular environment didn’t help. On the first page, the forum readers see in red text, “REMEMBER THIS IS AN OFFICIAL OCZ FORUM, NO LINKS OR DISCUSSION TO COMPETITORS PRODUCTS WILL BE TOLERATED.” The company certainly has the right to enforce that on their own site. But it doesn’t help them keep an open mind about unexpected results.

    Indilinx could soon release a firmware upgrade that fixes the demonstrated XP problems and all of this becomes irrelevant. Or some reliable, unobtrusive maintence plan (like some commenters advocated in the last article) could acceptably address the weaknesses as well. But I have to think that the OCZ’s credibility has been damaged, which wouldn’t have happenned (to this degree) if they didn’t mention the TR review in their fourm. And it’s not like this is the first weakness of OCZ SSDs that has exposed by a neutral review site; it’s at least their third time.

    And I won’t buy their argument that XP is irrelevant. Several OCZ fourm members, including the moderator himself (who remarked about the review, “LOL what a bunch of crap” and “Who knows what unforseen bottlenecks that system might have.. eeezaly.”) have XP and pre-ICH10 southbridges. The moderator’s profile can’t be that outdated, because he lists his three Vertex drives, which were released recently in March.

    In some ways I’m happy about their involvement, and I think in the end, consumers benefited from the follow-up TR articles. Geoff’s first article had some unexpected results that few, if any, could explain. But I also believe that we could have gotten an follow-up article, using a tactic that was open-minded and more respectful.

      • jwilliams
      • 12 years ago

      wibeasley:

      I am not sure why you say I am affiliated with OCZ. I have not posted a word about OCZ until this post.

      And I am still waiting for an explanation from Tech Report as to why every single SSD they tested booted more slowly on their system than two hard drives.

        • wibeasley
        • 12 years ago

        I apologize if I was wrong; I assumed you were the same as the OCZ forum member ‘jwilliams315’ who started that thread I linked. I didn’t mean to imply that you were an OCZ employee.

        I agree the boot time results are puzzling, and not still explained (at least in a way that I can understand them). Some commenters have given plausible explanations, but I’d still like to hear something from TR, and ideally replicated by other review sites. As the comments in the previous article/podcast mentioned, there are a variety of ways to define ‘boot time’. I like that the different sites have different ways of defining boot times, but it makes it difficult to rule out mistakes for any single site.

        I imagine the upcoming Vista review will include boot times, and hopefully a fuller explanation of the boot results (regardless if the Vista patterns are consistent with the XP patterns). But if they’re not included, I think you have a valid reason to ask for Geoff’s best guess why the used SSD drives appear to boot more slowly than expected.

        Edit: l[<"I have not posted a word about OCZ until this post"<]l I was referring to your podcast comments (#3, #9, #12, & #14).

          • derFunkenstein
          • 12 years ago

          No, the way they’re both banging the drum about the boot times, I think you’re right.

          And since that question has been answered, I assume we’re ready to move onto something else.

      • nerdrage
      • 12 years ago

      q[

        • derFunkenstein
        • 12 years ago

        What’s REALLY amusing to me is that TR has, time and time again, given OCZ’s products not just a fair shake but glowing reviews. The one time they note a flaw in a product, the OCZ fanboys go apeshit.

          • CasbahBoy
          • 12 years ago

          Seriously. Their (OCZ/PCPower&Cooling) power supplies have been rated as some of the absolute best. Maybe it is because I like tinkering with computer hardware generally (or maybe because I don’t invest too much of myself into the products that I purchase), but I own a Vertex and I’m loving the attention and criticism that these things have been getting because I look forward to a time when SSDs can be used and performance maintained without any user intervention or heighted risk of data loss.

        • Vaughn
        • 12 years ago

        lmao dude its not 1999 its 2009!

        You need to update your info and stop living in the past.

        You’re as bad as the guys that still think Ati’s drivers are shiet and Nvidia is still the gold standard!

        Thank you for a good laugh this evening!

          • wibeasley
          • 12 years ago

          MadManOriginal, would you consider this guy frothy, or just foamy?

            • MadManOriginal
            • 12 years ago

            Halfcaf latte with skim foam?

      • MadManOriginal
      • 12 years ago

      Ha, that thread is funny. You’d think a company would take reviews of which they are aware seriously and work with the reviewer not to fudge results or just make them look better but to at least see what’s going on. Instead they bash the review and the fanboys get all frothy. Pretty childish really.

        • wibeasley
        • 12 years ago

        Fortunately, they’re not all fanboys. Many of them appear to have a nice balance of skepticism and open-mindedness, and contributed their opinions of how their SSD use is different from the scenario modeled by Geoff (e.g., Chloiber). It’s helped broaden my perspective.

        And judging from the second page of this article, it appears that actual OCZ employees are open to the idea that manageable weakness still exist. I wonder if the actual OCZ employees wish that those forum members hadn’t started such an ‘enthusiastic’ thread, and instead just let them work out the issues with less publicity.

          • MadManOriginal
          • 12 years ago

          I didn’t say they’re all fanboys, just that the fanboys get frothy 🙂 At the same time reading some of the comments from the moderators which can be considered semi-official although maybe they aren’t OCZ employees they ought to be vetted, and ones from OCZ employees in that thread makes me shake my head.

            • derFunkenstein
            • 12 years ago

            Yeah, the third post of the thread is a typical mod forgetting they’re not God; they’re a representative of the company.

    • JoeKiller
    • 12 years ago

    Great article and investigation afterward. This is what the TR is all about. Keep on bringing the interesting tests.

    • _Sigma
    • 12 years ago

    Just so I’m clear, the current thinking is that the offset of WinXp is what is causing the poor Indilinx results?

    Perhaps I missed this…?

    • FireGryphon
    • 12 years ago

    Won’t constantly running the TRIM command (or some other similar OS-level command) unnecessarily write to the drive and cause it to deteriorate faster?

    Also, I don’t understand why SSDs have to take extra steps to read through blocks that are already written even if they’re marked as deleted. Why is that so?

      • Freon
      • 12 years ago

      Good points, and it has been brought up.

      I guess I wouldn’t have a huge problem with TRIM if I can schedule it to run every Sunday at 3AM or whatever. And if it didn’t cause data corruption, of course… Periodic maintenance that would supplant defragging is not that big of a deal. Definitely not a deal breaker IMHO.

      • CasbahBoy
      • 12 years ago

      I was under the impression that all blocks that are affected by TRIM would require a write erase prior to being written to again anyway, so you’re just performing the write erase before an actual time sensitive write is pending.

      • DrDillyBar
      • 12 years ago

      May be good to revisit this little info repository…
      §[<http://blogs.msdn.com/e7/archive/2009/05/05/support-and-q-a-for-solid-state-drives-and.aspx<]§

      • SNM
      • 12 years ago

      The TRIM command moves the erasure of a block from write (when it’s needed) to deletion (when it’s marked as empty and deleted by the filesystem). It doesn’t add any extra erases compared to waiting to perform the actual deletion at a later time.

      I don’t think you quite understand how the issue occurs, though. You don’t need to read through blocks that are already deleted. You DO have to delete things in 512k chunks, though, (and you can’t just overwrite) and that means sometimes you have to read data, delete the sector, and then write back the data as you write in new data.

    • clone
    • 12 years ago

    … man am I glad I waited and am still waiting before buying an SSD.

    • glynor
    • 12 years ago

    Great job as always, Geoff.

    I stayed out of the discussions on this on the original article and podcast article, but now that this additional information is available… I really feel that it is completely irrelevant if the Barefoot drives perform well when you run them on Windows Vista 64-bit, with an ICH10R chipset, with a specific (non-factory default) firmware revision, after scheduling and running a beta (and possibly data corrupting) third-party utility, and with all the other ands, ifs, and buts.

    A hard drive is a hard drive is a hard drive. It should work without needing to “fiddle” with it and endlessly monitor performance, and work with the prevalent, existing consumer operating systems.

    If OCZ and Indilinx specifically advertised this drive as only compatible with ICH10R and Windows 7 (or perhaps Vista 64-bit, though I too remain skeptical), then maybe I’d feel differently. As it is, though, any normal consumer with Windows XP and a relatively modern computer with SATA 3Gbps ports, would and should just expect the device to work, and it clearly does not.

    So, thanks again for all of your extremely hard work in detailing this issue. I look forward to seeing the results with the new rig, but mostly only for informational purposes. Even if it performs perfectly with Vista 64-bit and the ICH10R, that doesn’t change the fact that the device is fundamentally broken (or at least falsely advertised).

      • nerdrage
      • 12 years ago

      My thoughts exactly. There should be no need to update drivers, BIOS, firmware, chipset, or operating system just to get a hard drive to work properly. Not to mention that many new laptops ship with an older version of the ICH (my wife’s laptop shipped with ICH8M in January), and all the other non-Intel chipsets in use. In 2009, a hard drive should be plug and play, period.

      To me this screams “not ready for prime time”.

        • Chloiber
        • 12 years ago

        I don’t know any SSD which didn’t have a FW-Upgrade.

        SSDs aren’t HDDs. There is a mighty controller inside which needs updates from time to time – at least in the first few month after release. Even enterprise-SSDs like the MTRONs got them.
        And…SSDs are a new technology. Old drivers/components/OSes don’t work properly with this new form of “HDD”.

        I agree that this is the way that it “should” be, but this isn’t the way it is and will be in the near future. Consumer – SSDs are still “beta” – every single one. Of course with Indilinx-Drives you got more problems at the moment, I admit that. But if you wan’t a “new” SSD which you can just plug in, do nothing and be happy for the next years you have to wait for some more years.

        And by the way: my UD is again in my Notebook with ICH8M and Win7 running flawlessly 😉

          • MadManOriginal
          • 12 years ago

          If all consumer SSDs are still ‘beta’ then none should be recommended. Getting to try beta software (or hardware – rare) for free is one thing but paying SSD prices to be a beta tester?

            • Chloiber
            • 12 years ago

            Well but that’s the way it is 😉

            Intel had their 80% bug, are still having performance issues with their X25-E, performance drops with X25-M switching between “high IOPS” and “high Datarate”,…

            JMicron got stuttering…

            I won’t start talking about Indilinx, I think by the time you know what bugs there are 😉

            I’m just saying: nearly every SSD got some bugs and you will have to tweak some things. There is no Plug and Play.

          • glynor
          • 12 years ago

          If it looks like a hard drive, and it quacks like a hard drive, then it IS a hard drive.

          If the companies selling them want to market them like a hard drive: with the same connector, filling the same product space in the market, with the same form factor, and want to compare read/write speeds directly to mechanical hard drives; then they need to be held to the same reliability and ease-of-use standards as commodity magnetic hard drives.

          Currently, in this regard, most SSDs fail on the margins. With these kinds of results, though, it seems that these Indilinx drives fail in a much more fundamental way, and in a much less forgivable way.

          I don’t dispute that SSDs might be right for some people currently. And they likely will be fine for me as the technology matures. But you can’t fault TR for pointing out the, quite obvious, shortcomings that they encountered when working with the product, and claim “but they did it wrong”.

          A consumer-marketed hard drive “product” absolutely still needs to be able to work with the type of system tested here. Period. Anything else is excuses and fiddling in the margins.

            • UberGerbil
            • 12 years ago

            You’re aware that hard drives sometimes get updated firmware also? Particularly high performance hard drives aimed at the enthusiast segment of the market? The Raptor in particular went through a series of firmware revisions, each of which increased performance (especially as OSes evolved to support TCQ and then NCQ).
            §[<http://www.storagereview.com/articles/200512/WD740GD-00FLC0_1.html<]§ However they weren't field-upgradable: to get the improved firmware, you either had to send your drive back to Western Digital or simply buy a new drive. At least with SSDs you mostly can update them in place. I'm not trying to excuse the SSD vendors: I agree their products need to work with existing hardware, including old hardware, and it should deliver claimed performance out of the box. Clearly, there's a lot of R&D left to do -- more for some controllers than others -- and the hype may have overtaken the reality. But it appears they are doing the work. And if hard drives are your golden standard, you need to be aware they've had their teething pains also. Such is the life of the enthusiast on the bleeding edge. And kudos to TR for pointing it out.

          • Freon
          • 12 years ago

          “…needs updates from time to time…”

          *record scratch*

          Well, I hope not. I hope this is just growing pains and that this entire class of issue is really put to bed in the coming months. I’m definitely not confident in buying an SSD if this is the prevailing attitude.

            • Chloiber
            • 12 years ago

            I don’t know of ANY controller which doesn’t get updates – at least in the first several month after release.

            • Freon
            • 12 years ago

            I’ve never had to update a hard drive firmware. I’ve never even heard of people having to do that until Seagate-gate a few months back. Maybe I live in a cave and this is really widespread…

            • Usacomp2k3
            • 12 years ago

            Me neither. In fact, I don’t think I’ve ever updated the firmware on anything before, except for my Zune or a cell phone.

      • eitje
      • 12 years ago
        • glynor
        • 12 years ago

        (In reply to #58 as well)

        Oh, if the problems were limited to Firmware fixes only, I might completely agree, and that is certainly why I waited for this updated article before commenting.

        If the issues were just a “they didn’t use the latest firmware to test” issue, then I’d say that the issue could drop with a quick “Sorry I didn’t use the latest firmware in testing, but now it is fixed” note from Geoff (with a “why are they still shipping broken firmware in new drives” mumble on the side).

        That isn’t currently the case at all. As the new tests show, the Firmware update doesn’t fix the issues here. Now, some users (hopefully well informed) might assume that a future firmware update will resolve all issues, but I’m not going to do that. If you can’t fix it now, then it is broken now, and they should report that as such.

        My point is, and was, that the cumulative effect of needing lots of ands, ifs, and buts (including but not limited to firmware updates) to even have a chance of having a well-performing drive, to me, makes some of those “ands, ifs, and buts” irrelevant side-notes to the main thrust of the review which is:

        These drives don’t perform as well as the competition in many tested configurations, due to controller issues with a used-drive state, and are in effect, broken.

        That said… I would like to see some comment, in one way or another, from Geoff on the raised boot-time issue… Is the Intel 25-M booting slower than many of these mechanical drives? I know that SSDs often aren’t quite the boot-time panacea that they are made out to be, but they do often rate at least somewhat faster, correct?

          • bats
          • 12 years ago

          g{

    • GFC
    • 12 years ago

    I’d say some commenters need to keep quiet, theese ppl work their ass off, and if some guys at taiwan screw up with the controller – that’s not their fault.

      • indeego
      • 12 years ago

      So who can we blame for this malcontent? Heads will ROLLg{

      • pmonti80
      • 12 years ago

      Just so you know, anandtech also tested all the SSD’s in an used state and didn’t find such a big difference. But maybe the next test TR is doing will shed some needed light into these problems the found. It will be interesting to see what happens whatever are the final results.

        • Freon
        • 12 years ago

        “Used state” seems open to wide interpretation. As TR covered in the podcast, artiles don’t always get into enough detail about how they test so as to allow others to reproduce their results.

        Good science should be reproducible by anyone with the proper equipment and the ability to follow a written procedure.

          • pmonti80
          • 12 years ago

          This line could be taken exactly from anand and that’s the type of attitude they have towards testing at anandtech. I’m not blaming anyone, but saying anand is unscientific is not exactly true.
          The reason I usually follow Anandtech and TR is that they try to be as scientific as possible while some other sites don’t truly understand what they are doing (SSD’s are the prime example). Sometimes I disagree with one or another and this time I’m just not totally convinced by Techreport’s explanations, but we seam to be slowly getting there.

            • Freon
            • 12 years ago

            Yeah, I didn’t mean to imply Anandtech is unscientific. I trust them as well. It’s just that there doesn’t seem to be an agreed standard for preparing a drive as “used” so it certainly is not a universal term in this context. If TR’s “used” state is more truly “used” than Anand’s it could explain why TR shows poorer performance.

            For now it might be better than there is no agreed standard, but better documentation on procedure by all would be helpful in tracking down possibly conditions causing performance quirks. Again, the point here is reproducibility.

        • SNM
        • 12 years ago

        I thought Anand’s differences were pretty huge as well (and, just as importantly, the same order). But if they weren’t similar enough, hasn’t Anand moved over to Vista for his standard test machine?

          • derFunkenstein
          • 12 years ago

          Yes, you are correct, yet teh OCZ forum fanbase insists on slaying TR and TR alone.

            • pmonti80
            • 12 years ago

            Never been on an OCZ fanboy, mainly because I don’t have any OCZ product or any SSD for that matter. And to prove it I will tell you that I find the results of the OCZ Apex a little too good here.

            The thing is that anand had +/- the same kind of performance hit on both SSD’s. Also, all the results you see were with the used state, and in almost every one of them both Intel SSD and Vertex beat hands down a WD VelociRaptor.

            I’m not bashing TechReport, mainly because they have said that there will be a 3rd article. I will wait to praise or criticize them on the 3rd article.

      • derFunkenstein
      • 12 years ago

      Continuing to sell them after the problem has been recognized is the fault of everyone down the chain.

      • morphine
      • 12 years ago

      Funny… I worked my ass off to buy an SSD and it doesn’t work as supposed to because, you know, someone happened to be incompentent designing its controller and I gave him/her my money. Oh, the humanity!

      • CasbahBoy
      • 12 years ago

      Awesome, so I should be happy with the dead pixels on my LCD at work because some people worked their asses off to make the displays, and I should be happy with a transmission that fails at 10,000 miles in my Ford because some people worked their asses off to design and build it.

      A flawed product is a flawed product regardless of the time and effort gone into producing it.

    • Igor_Kavinski
    • 12 years ago

    Hate to say this but these benchmark results are irrelevant because the current Vertex firmware version is V1.10. Maybe you guys should have redone the benchmarks before posting the article?

      • green
      • 12 years ago

      that’s perfect! why didn’t i think of that before…
      release firmware revisions regularly enough that published results are invalidated for using an outdated revision!

      i gotta tell the boss…

      • Chloiber
      • 12 years ago

      1370 = 1.10

    • Faiakes
    • 12 years ago

    So during the first review and/or the second one, was the Vertex ever tested with the recommended 64 sector alignment?

    If not, will it be tested as such in the upcoming review?

      • Chloiber
      • 12 years ago

      It will, because Vista doesn’t align the partition like XP. Vista (afaik) aligns with 1024KB – that’s okay. You won’t see big differences between 64k or 1024k (if any).

    • S_D
    • 12 years ago

    ‘Props’ to the TR team for picking up on the ‘criticism’ of the previous article, and trying to isolate the cause(s) for the results. Your work here is really appreciate (as a potential future Vertex customer myself), and I look forward to your new tests results with great anticipation.

    • odizzido
    • 12 years ago

    Excellent article(s) guys. I had no idea these sorts of problems existed until I read about them here.

    • MadManOriginal
    • 12 years ago

    I was going to say something less than flattering about going through all this retesting on the same platform until I read the last paragraph. I hope to see those results soon and it’s not clear but I hope you’re going to use new hardware in addition to a new OS. One of the guys in the last article comments said that there may be problems with ICH7 despite it supposedly being identical to later ICH. Doing both new OS+old hardware and new OS+newer hardware would be best to cover all single variable chagnes 🙂

      • Meadows
      • 12 years ago

      Someone said a firmware- or BIOS related thing about /[

        • indeego
        • 12 years ago
        • MadManOriginal
        • 12 years ago

        Yeah it’s in the OCZ thread that’s in post #33. Supposedly everything pre-ICH10 using AHCI/RAID setting has a BIOS that extends boot times greatly for SSDs. The fix is to hack the motherboard BIOS with newer AHCI/RAID firmware :rolleyes: I don’t know if there’s a simple way to update the AHCI/RAID firmware otherwise. I can see doing some tweaks for people who don’t mind that but hacking a motherboard BIOS is just so over the top…

      • pmonti80
      • 12 years ago

      Agreed……

      • barich
      • 12 years ago

      If I read the same comment you did, that was related to the RAID ROM and only affected boot time.

    • pmonti80
    • 12 years ago

    I will reserve my judgement to when the next test will be done.

    • DrDillyBar
    • 12 years ago

    *high 5 to TR*
    I’ve got ICH7 and Win7. Yay me.

    • Meadows
    • 12 years ago

    g{

      • Dissonance
      • 12 years ago

      Fixed.

        • Palek
        • 12 years ago

        Same paragraph:

        q[

Pin It on Pinterest

Share This