Samsung’s 860 EVO 1 TB SSD reviewed

Samsung’s 850 EVO has reigned supreme in the SATA SSD market for a long time. It earned our recommendation when it launched all the way back in 2014 and has been the 800-pound gorilla of client SSDs ever since. Samsung was the first to produce a non-planar TLC NAND drive, and the 850 EVO’s success has affected the shape of the entire SSD market since. Pick up a random SSD and peer inside, and the odds are high that you’ll find 3D TLC NAND of some kind staring back at you.

Non-planar NAND technologies–whether “3D,” “BiCS,” or “V-NAND”–have become the de facto standard for mainstream solid-state storage. And that’s not likely to change soon, unless more exotic technologies coalesce from the vapor. The only credible threat comes from quad-level-cell NAND. For now, QLC products seem to be targeted towards enterprise-y “write once, read many” workloads, where the technology’s inevitably poorer write performance and endurance can go unnoticed. But perhaps it’s only a matter of time before some plucky manufacturer decides that the cost savings outweigh the potential consumer backlash and forces it into the client space. I expect much wailing and gnashing of teeth in the comments when it happens, so start composing your diatribes today.

Anyway, the name of the game is still non-planar TLC for now, and as we discussed in our 860 Pro review, Samsung has spent the last few years making generational refinements to its V-NAND. Meet the 860 EVO and the 64 layers of TLC V-NAND goodness in each of its flash packages.

Samsung 860 EVO
Capacity Max sequential (MB/s) Max random (IOps)
Read Write Read Write
250 GB 550 520 98K 90K
500 GB 550 520 98K 90K
1 TB 550 520 98K 90K
2 TB 550 520 98K 90K
4 TB 550 520 98K 90K

Now, this drive has been out for most of this year, but the one-two punch of short initial review sample supply and our recently-finished efforts to modernize our test rig knocked this write-up pretty far down the stretch. Our apologies for the delay. Samsung hooked us up with the 1-TB version of the drive, which turns out to be the middle child. The lineup includes the standard 250 GB and 500 GB capacities, and gargantuan 2-TB and 4-TB variants are also available. The 850 EVO series also got 2-TB and 4-TB flavors, but not until well after that drive launched.

The difference between the 860 EVO and Pro, as usual, is TLC V-NAND instead of MLC. Theoretically, that should mean reduced performance, but Samsung’s Intelligent TurboWrite caching scheme will ensure that the EVO operates at near-SATA-6-Gbps limits for most workloads in spite of its extra-bit-per-cell handicap.

Once its shell is shucked, the 860 EVO 1 TB looks almost comically small. Samsung crams all that TLC V-NAND into only two packages: one on either side of the tiny PCB. One of those packages shares elbow room with 1 GB of LPDDR4 RAM and the same MJX controller used in the 860 Pro.

Much like the  860 Pro, the 860 EVO is chock-full of helpful encryption capabilities and comes with a five-year warranty. Unlike the Pro 1 TB, however, the EVO 1 TB gets an endurance rating of 600 TBW, merely half of its higher-end sibling’s 1200-TBW specification. That’s still more than enough headroom to satisfy any ordinary usage.

Samsung’s suggested price at launch was $330 for the 1 TB version, but we’re currently experiencing something of a renaissance for cheap SSDs. The drive is available for $218 at Amazon, marked down a third from Samsung’s projected ask. If the drive’s speeds are anything like the 850 EVO’s, that will make for some incredible performance per dollar. Let’s see how the new guy stands up to our test suite.

 

IOMeter — Sequential and random performance

IOMeter fuels much of our latest storage test suite, including our sequential and random I/O tests. These tests are run across the full capacity of the drive at two queue depths. The QD1 tests simulate a single thread, while the QD4 results emulate a more demanding desktop workload. For perspective, 87% of the requests in our old DriveBench 2.0 trace of real-world desktop activity have a queue depth of four or less. Clicking the buttons below the graphs switches between results charted at the different queue depths. Our sequential tests use a relatively large 128-KB block size.



All’s well so far. The 860 EVO reads a hair faster than its predecessor. It also writes slightly faster at QD1, but the 850 EVO makes an appreciable gain at QD4 while the 860 EVO gains almost nothing.



Random read response times are close to dead even between the two EVOs. The 860’s random writes, however, are quite a bit snappier at QD4.

It’s more a less a wash between the two generations of EVO series drives so far. But that’s by no means a bad thing–there’s not much ground left to gain for SATA drives. Let’s see how the 860 EVO does in our sustained and scaling tests.

 

Sustained and scaling I/O rates

Our sustained IOMeter test hammers drives with 4-KB random writes for 30 minutes straight. It uses a queue depth of 32, a setting that should result in higher speeds that saturate each drive’s overprovisioned area more quickly. This lengthy—and heavy—workload isn’t indicative of typical PC use, but it provides a sense of how the drives react when they’re pushed to the brink.

We’re reporting IOPS rather than response times for these tests. Click the buttons below the graph to switch between SSDs.


The two EVOs’ peak and steady-state write rates are very nearly identical, but it looks like the 860 EVO holds onto its maximum for substantially longer. The 850 EVO ramps down sometime after 150 seconds have passed, while the 860 EVO chugs along for an additional 100 seconds before declining. 

As suspected, the drives are neck and neck for throughput, but the 860 EVO’s longer high-performance burst is a pleasing upgrade.

Our final IOMeter test examines performance scaling across a broad range of queue depths. We ramp all the way up to a queue depth of 128. Don’t expect AHCI-based drives to scale past 32, though—that’s the maximum depth of their native command queues.

For this test, we use a database access pattern comprising 66% reads and 33% writes, all of which are random. The test runs after 30 minutes of continuous random writes that put the drive in a simulated used state. Click the buttons below the graph to switch between the different drives. Note that each drive uses a different scale for IOPS to allow us to view the shape of its curves.


The 860 EVO’s performance ramps up quickly, nearing maximum speeds even at QD4. The 850 EVO scaled more slowly, only approaching its maximum at QD16. Let’s look at the drives plotted against each other.


The curves lose some of the their impact when the graph is adjusted to fit the 970 EVO’s NVMe-endowed scaling prowess, but both of the lesser EVOs exhibit noticeable scaling for AHCI drives.  The 860 EVO’s throughput more than doubles from QD1 to QD4.

The 860 EVO continues to come out slightly ahead of its forebear, which bodes well for its place in our overall rankings. Now it’s time to put aside IOMeter and transition into real-world testing.

 

TR RoboBench — Real-world transfers

RoboBench trades synthetic tests with random data for real-world transfers with a range of file types. Developed by our in-house coder, Bruno “morphine” Ferreira, this benchmark relies on the multi-threaded robocopy command build into Windows. We copy files to and from a wicked-fast RAM disk to measure read and write performance. We also cut the RAM disk out of the loop for a copy test that transfers the files to a different location on the SSD.

Robocopy uses eight threads by default, and we’ve also run it with a single thread. Our results are split between two file sets, whose vital statistics are detailed below. The compressibility percentage is based on the size of the file set after it’s been crunched by 7-Zip.

  Number of files Average file size Total size Compressibility
Media 459 21.4MB 9.58GB 0.8%
Work 84,652 48.0KB 3.87GB 59%

RoboBench’s write and copy tests run after the drives have been put into a simulated used state with 30 minutes of 4KB random writes. The pre-conditioning process is scripted, as is the rest of the test, ensuring that drives have the same amount of time to recover.

The media set is made up of large movie files, high-bitrate MP3s, and 18-megapixel RAW and JPG images. There are only a few hundred files in total, and the data set isn’t amenable to compression. The work set comprises loads of TR files, including documents, spreadsheets, and web-optimized images. It also includes a stack of programming-related files associated with our old Mozilla compiling test and the Visual Studio test on the next page. The average file size is measured in kilobytes rather than megabytes, and the files are mostly compressible.

Let’s take a look at the media set first. The buttons switch between read, write, and copy results.



It’s close to even between the 850 EVO and 860 EVO in the media set. The write tests offer the most differentiation, but unfortunately it’s in favor of the older drive. We’re talking a 5-8% difference though, so it’s not a big deal. Maybe the work set will change the outlook.



This time we can call it a tie. The two drives trade blows across read, write, and copy tests.

Overall, the 850 EVO and 860 EVO behave largely the same in RoboBench, as they did with IOMeter. But if it ain’t broke, don’t fix it. Our last page of tests will see how it fares when it’s expected to load up an operating system.

 

Boot times

Until now, all of our tests have been conducted with the SSDs connected as secondary storage. This next batch uses them as system drives.

We’ll start with boot times measured two ways. The bare test depicts the time between hitting the power button and reaching the Windows desktop, while the loaded test adds the time needed to load four applications—Avidemux, LibreOffice, GIMP, and Visual Studio—automatically from the startup folder. These tests cover the entire boot process, including drive initialization.

The 860 EVO will have you booted up before you’ve finished putting your face on for Windows Hello. Of course, so will the 850 EVO. 

Load times

Next, we’ll tackle load times with two sets of tests. The first group focuses on the time required to load larger files in a collection of desktop applications. We open a 790-MB 4K video in Avidemux, a 30-MB spreadsheet in LibreOffice, and a 523-MB image file in the GIMP. In the Visual Studio Express test, we open a 159-MB project containing source code for Microsoft’s PowerShell.

Load times for the first three programs are recorded using PassMark AppTimer. AppTimer’s load completion detection doesn’t play nice with Visual Studio, so we’re still using a stopwatch for that one.

Yet again, the two EVO drives yield remarkably similar results. Visual Studio is more than half a second kinder to the newer drive, though.

If you can’t spare the 0.3-or-less extra seconds it takes the 860 EVO to load games, you better stick with the 850 EVO. Either drive will serve your Steam library well.

That’s all of our tests. The 860 EVO gave us no surprises while serving as a boot drive. Our test methods are on the next page, and our conclusions on the following.

 

Test notes and methods

Here are the essential details for all the drives we tested:

  Interface Flash controller NAND
Adata XPG SX8200 480GB PCIe Gen3 x4 Silicon Motion SM2262 64-layer Micron 3D TLC
Crucial MX500 500GB SATA 6Gbps Silicon Motion SM2258 64-layer Micron 3D TLC
Intel X25-M G2 160GB SATA 3Gbps Intel PC29AS21BA0 34-nm Intel MLC
Samsung 850 EVO 1TB SATA 6Gbps Samsung MEX 32-layer Samsung TLC
Samsung 860 EVO 1TB SATA 6Gbps Samsung MJX 64-layer Samsung TLC
Samsung 970 EVO 1TB PCIe Gen3 x4 Samsung Phoenix 64-layer Samsung TLC
Toshiba RC100 PCIe Gen3 x2 Toshiba 64-layer Toshiba BiCS TLC

The SATA SSDs were connected to the motherboard’s Z270 chipset. The PCIe drives were connected via one of the motherboard’s M.2 slots, which also draw their lanes from the Z270 chipset.

We used the following system for testing:

Processor Intel Core i7-6700K
Motherboard Gigabyte Aorus Z270X-Gaming 5
Firmware F10B
Memory size 16 GB (2 DIMMs)
Memory type Corsair Vengeance LPX DDR4 at 2133 MT/s
Memory timings 15-17-17-35
System drive Corsair Force LS 240GB with S8FM07.9 firmware
Power supply Rosewill Fortress 550 W
Operating system Windows 10 x64 1803

Thanks to Gigabyte for providing the system’s motherboard, to Intel for the CPU, to Corsair for the memory and system drive, and to Rosewill for the PSU. And thanks to the drive makers for supplying the rest of the SSDs.

We used the following versions of our test applications:

Some further notes on our test methods:

  • To ensure consistent and repeatable results, the SSDs were secure-erased before every component of our test suite. For the IOMeter database, RoboBench write, and RoboBench copy tests, the drives were put in a simulated used state that better exposes long-term performance characteristics. Those tests are all scripted, ensuring an even playing field that gives the drives the same amount of time to recover from the initial used state.

  • We run virtually all our tests three times and report the median of the results. Our sustained IOMeter test is run a second time to verify the results of the first test and additional times only if necessary. The sustained test runs for 30 minutes continuously, so it already samples performance over a long period.

  • Steps have been taken to ensure the CPU’s power-saving features don’t taint any of our results. All of the CPU’s low-power states have been disabled, effectively pegging the frequency at 4.0 GHz. Transitioning between power states can affect the performance of storage benchmarks, especially when dealing with short burst transfers.

The test systems’ Windows desktop was set at 1920×1200 at 60 Hz. Most of the tests and methods we employed are publicly available and reproducible. If you have questions about our methods, hit our forums to talk with us about them.

 

Conclusions

The 860 EVO 1 TB gave us the solid mainstream performance we’ve come to expect out of Samsung’s EVO line. We distill the overall performance rating using an older SATA SSD as a baseline. To compare each drive, we then take the geometric mean of a basket of results from our test suite.

The 860 EVO 1 TB is the fastest SATA drive thus far in our burgeoning new test set. As with the 860 Pro, Samsung’s goal for the 860 EVO was to transition its mainstream products onto its latest mass-produced NAND technologies without introducing performance regressions. Clearly, the firm has done so with resounding success.

The intro had some spoilers for this next part, so you may already suspect how the 860 EVO 1 TB will fall on our value scatter plot. In the graph below, the most compelling position is toward the upper left corner, where the price per gigabyte is low and performance is high.

A terabyte of solid-state storage just doesn’t yield a retailer as many dollars as it used to. It’s a great time to buy an SSD in any market segment. Amazon’s $218 price tag for the 860 EVO translates to a tempting 22 cents per gigabyte. I remember buying an 850 EVO 500 GB for $150 some years ago (or 30 cents per gig) and thinking it was a steal. $0.22 per gig is a price that until recently was reserved for truly bargain-basement drives with shorter warranties, lower performance, and no encryption features.

The 860 EVO would likely have gotten our recommendation even without today’s pricing moves in play, but right now it’s really a no-brainer. Though the novelty has faded from the technologies the drive is built on, the 860 EVO still offers a winning blend of strong performance, high endurance, and quality-of-life features like 256-bit hardware encryption acceleration in its controller. The only thing that might hold a prospective buyer back is the call of cheaper NVMe drives like Adata’s SX8200. But if you’ve been on the prowl for a meaty chunk of SATA space, the time is ripe. You can’t go wrong with an 860 EVO 1 TB.

Comments closed
    • kamikaziechameleon
    • 1 year ago

    Any thoughts on how this compares performance wise to the pro equivalent?

    • TheMonkeyKing
    • 1 year ago

    Price on Amazon dropped below the MX500 – 1TB to $179.99, $189 for the MX500.

    8/27/18

    • Jigar
    • 1 year ago

    My 120GB 850 EVO does boot time in 24 seconds after 2 years of use, guess i should be happy.

    • LoneWolf15
    • 1 year ago

    Good article –but as others have perhaps hinted, I think as much attention should have been paid to comparing the Crucial MX500 as to the earlier 850 Evo.

    Last week, the MX500 1TB was $177 on Amazon, which is dirt cheap. It’s hard to justify $40 price difference unless there’s significantly higher longevity, better reliability, significantly lower power usage, or significantly faster speed on the 860 EVO. The MX500 1TB is now back to $193, but that’s still a $25 difference. The MX500 supports hardware encryption, and other important features that the 860 EVO does, so it’s tough to say the 860 EVO has a commanding lead.

      • strangerguy
      • 1 year ago

      Now with the HP EX920 1TB NVMe hitting $250 (on TR’s current frontpage no less), nobody in the right mind should be considering any SATA 1TB drive for $200+ ever.

        • LoneWolf15
        • 1 year ago

        It depends on what their system interface supports. I don’t have an M.2 and while I can use an NVMe drive, that’s only if it’s in 2.5″ form factor (my ThinkPad T460p will support it over a SATA-Express compatible SATA inferface; such drives are extremely rare).

        Longer-length M.2 drives would be to long to fit in my notebook with an adapter, and at 1TB sizes, that’s all one is going to find.

    • shank15217
    • 1 year ago

    This means they could easily make a 8TB SSD drive for consumers that no one can afford.

    • barich
    • 1 year ago

    It’s remarkable how competitive the X25-M G2 still is for read-biased desktop usage (boot and application load times). It’s obviously well behind in sequential reads and especially writes, but as a boot/application drive I doubt you’d notice obvious performance problems other than when installing software and updates.

      • Chrispy_
      • 1 year ago

      It just confirms how desktop usage is highly dependent on 4KQD1 performance and that NVMe marketing numbers really aren’t as relevant as the salesmen would like you to believe.

      The first thing I look at in any SSD review is 4KQD1 performance. Anything under 5K IOPS and the drive is junk to be avoided because [i<]bad[/i<] drives from half a decade ago can manage that. When you see the "up to 200K IOPS" rubbish it's totally irrelevant because you'll be lucky if it ever reaches 20K in a typical consumer workload. The class-leading SATA controller on the market for QD1 performance is still the MX500 and that tops out at a 4K data rate of just 45MB/s, which is less than 12K IOPS - a drive that the marketing goons advertise as a 95K IOPS performer* *95K IOPS achieved in a special test we wrote ourselves using best-case scenario data and at max queue depth in a brand-new, virgin state on a Core i9 clocked at 7GHz running a custom OS we wrote that only understands our SSD and can't do anything except run this one test utility. Value shown is the peak outlier result from 4000 test runs. YMMV E&OE.

        • HERETIC
        • 1 year ago

        My line in the sand is 8K IOPS.
        There’s no reason to go any lower today,now that everyone has caught up to Sammy…….

          • strangerguy
          • 1 year ago

          If there is a way to artificially cap and test IOPS at just 2000, I doubt it will affect desktop usage one bit anyway.

        • BurntMyBacon
        • 1 year ago

        I like your criteria. 4KQD1 is definitely the most common scenario for client type workloads. Though, I recall reading somewhere (Sorry, I can’t recall the link to back it up) that the average client reaches a max queue depth of 3 in typical usage and serious power users have been known to max at around 5 in rare circumstances. Barring specific applications or heavy server type workloads, these 4KQD32 or even higher QD numbers are less than meaningless as they only serve to distract from the valuable data points. It would be nice if manufactures posted 4KQD1 specs as they are always relevant. The max they should post on a consumer targeted drive is 4KQD4 as it is still somewhat significant.

    • limitedaccess
    • 1 year ago

    Can you go into more detail about why Techreport’s Shadows of Mordor load time results went from –

    [url<]https://techreport.com/review/33707/western-digital-black-1-tb-nvme-ssd-reviewed/5[/url<] to this [url<]https://techreport.com/review/33950/samsung-860-evo-1-tb-ssd-reviewed/5[/url<]

      • weaktoss
      • 1 year ago

      There’s no single definitive answer, but there are a lot of likely contributing factors.

      1) The older results were from a test image Geoff created sometime in early 2015. The latest results are from a test image I created last month. Shadow of Mordor has been patched in between, and some of those patches certainly had a performance focus.

      2) My save file is loading into a different area of the game than Geoff’s was.

      3) The difference in speed has forced me to alter the measurement point. During the loading screen, there’s some recap dialogue. If loading finishes before the spiel is over, a “press Esc to skip” prompt appears on screen. That’s the point I’m measuring now. Previously, the recap would always complete long before the game was finished loading, so the measure point was after the loading screen automatically went away and the scene actually rendered.

      4) Going from an i5-4690K to an i7-6700K makes a big difference.

    • Chrispy_
    • 1 year ago

    Samsung have done nice work on the all-important 4KQD1 performance. They’ve caught up to the MX500 which has been the obvious choice since it launched.

    If it was priced to compete with the MX500 I’d say it’ll sell like hotcakes but I am still personally boycotting Samsung for their 840 vanilla, 840EVO, PM841, and around a dozen different models of OEM drives that were used in Acer/Dell/Lenovo laptops as rebadges of the vanilla and EVO 840 models.

    Samsung drives are good, but they really failed hard with the 840 incident. Part of the blame is that they denied wrongdoing whilst their customers suffered for months. Part of the blame is treating people with failing drives like complete garbage. Part of the blame is refusing to recall defective drives and attempting to satisfy people with band-aid patches on faulty hardware, and the final nail in the coffin was that the firmware fixes were only ever released for the retail drives. If you had a laptop with an OEM model you were SOL and apart from throwing away your (once) expensive SSD your only other option is to run third-party band-aid utilities like diskfresh on a monthly basis.

    So anyone buying this drive will probably be fine and have a fast, reliable SATA SSD. Just know that if it turns out to be a turd, you’re going to get bent over and ****ed by Samsung – that’s their [i<]thing[/i<].

      • cygnus1
      • 1 year ago

      I had an 840 Evo as well. Twas also when I stopped buying Samsung drives and my 840 Evo became a scratch/non-critical stuff disk. There is always another brand that is comparable on price/performance. My buys have pretty much all been Crucial, with some Intel, SanDisk, and Mushkin thrown in.

      Did Samsung ever actually release a firmware fix for the vanilla 840? I thought they only ever fixed the Evo model.

        • barich
        • 1 year ago

        They issued a firmware update for the 840 like a year after they fixed the EVO.

        It doesn’t do anything on its own; it just allows you to run advanced performance optimization in Magician. The 840 vanilla’s performance degrades much more slowly than the EVO’s did, so running that a couple of times a year would be sufficient to keep it performing well.

          • Chrispy_
          • 1 year ago

          A year. Gee, how good of them for a 2-year old faulty product at that point.

          Also, AFAIK the ‘final’ EXT0DB6Q firmware for the 840 completely b0rked linux and was only a valid fix for Windows users.

          I have a box at work with 50+ useless 840-generation Samsung SSDs just tossed on the discard pile. I haven’t actually thrown it out because it represents around €15,000 of mistake I’m responsible for. In a corporate environment the delay and BS Samsung tried to feed consumers regarding a valid fix was completely unacceptable and we pulled them all as soon as it became apparent that Samsung was just going to let the problem fester rather than recall, admit their mistake or even apologize.

          I bought Samsung because they were the fastest and the most reputable drives at the time. A little bit like the Apple RDF, Samsung should be tarred and feathered but fickle consumers have ridiculously short memories, it would seem. Y’all hated on Sandforce and OCZ until they were bought out on the verge of bankruptcy – despite their product shortcomings being arguably less serious.

            • just brew it!
            • 1 year ago

            [quote<]Also, AFAIK the 'final' EXT0DB6Q firmware for the 840 completely b0rked linux and was only a valid fix for Windows users.[/quote<] The issue was allegedly (already) present in other Samsung 800-series models. With the EXT0DB6Q update they introduced the same bug into the 840 EVO. Linux kernels released after mid-2015 work around the issue (by disabling use of "queued TRIM" when a Samsung 800-series drive is detected). Windows was unaffected because it does not use queued TRIM.

      • BurntMyBacon
      • 1 year ago

      Point of interest:
      [quote=”Chrispy_”<]... and the final nail in the coffin was that the firmware fixes were only ever released for the retail drives. If you had a laptop with an OEM model you were SOL ...[/quote<] Samsung never released fixes directly for OEM drives. As far as I know, they gave the fixes directly to the laptop manufacturers so that they could QC them and left it to the manufactures to distribute. In any case, counter to your statement, my old work place had a bunch of Dells with OEM rebadged 840 EVOs that got the firmware fix (though it took far too long). Perhaps Samsung didn't give the fix to everyone, I don't have the inclination to look into it at this point. However, I find it more likely that the laptop manufacturers simply decided not to release the update (for all models?) or consumers simply never saw them. Having said all that, I still have to agree with your final assessment. Samsung was very late to act on information they had for a long time. They never [b<]properly[/b<] fixed the 840 (only the 840 EVO) as far as I'm aware. Considering how they've handled this and other debacles, I would only expect them to act once the issue becomes high profile and looks like it may lose them money in court. From the consumers standpoint, we should hope that any Samsung failure crashes hard and fast. [i<]Edited to reflect that the 840 did eventually get a fix, but it seems that it wasn't a full fix and had major drawbacks that prohibited Linux users from using it.[/i<]

      • Luminair
      • 1 year ago

      my 840 evo works fine. the firmware maintains itself when idle. the nand bug was that it loses data when disconnected from power for months. so just don’t use it for cold storage

        • Chrispy_
        • 1 year ago

        You’re oversimplifying the problem slightly.

        The NAND bug was far more serious than that:

        [list<][*<]It loses data when unpowered! [/*<][*<]OS and applications ran like garbage after a month or two even when powered. [/*<][*<]The band-aid software fix was Windows-exclusive and 5-6 months late. [/*<][*<]The firmware fix was over a year late! [/*<][*<]The fixes only applied to Windows. Linux users had to run Windows to update their faulty drives [/*<][*<]Both the software band-aid and the firmware fix affect drive endurance since they're constantly rewriting your NAND in the background, burning through the already low number of write/erase cycles of TLC NAND [/*<][*<]The final firmware has caused problems (data loss and corruption!) in Linux, because the fix breaks SATA TRIM support with Samsung's NAND re-write band-aid. The recommendation is to use old firmware and manually refresh your faulty drives using third-party tools. [/*<][*<]The final firmware fix exists only for the 840EVO and (eventually, after 26 months on the market) the vanilla 840. Users of the PM841 and at least half a dozen popular OEM variants are out of luck.[/*<][/list<] If you are a windows user with an 840EVO running Samsung Magician, you were the best-case scenario. All you actually suffered was sub-par performance for the first 5-6 months of the product lifecycle. If you are Linux user, Hackintosh user, running ESX or some other non-windows environment you were (still are) boned. If you didn't buy an EVO you were boned. The warranty ran out on most of the faulty drives long before Samsung even acknowledged the problem, if your drive is even one they officially acknowledged before. And the worst part about all of this? Samsung never admitted blame and sent paying customers on many a wild goose chase. It's in the same ballpark as Apple's 'you're holding it wrong' RDF.

          • barich
          • 1 year ago

          I agree with all of this. Frankly, they should have recalled and replaced all 840 and 840 EVO based products. They’re inherently defective and the firmware fixes at best worked around the problems and at worst created more serious ones.

      • just brew it!
      • 1 year ago

      Yup, Samsung is [i<]still[/i<] on my "avoid" list. Not just because of the bug itself (everyone has bugs from time to time), but also because of all the foot-dragging, obfuscation, and general bad behavior that followed. And the fact that it seems pretty clear in hindsight that the generation of planar TLC used in the 840 EVO was not quite ready for prime time, yet they launched a product based on it anyway. With solidly performing drives available at reasonable prices from other vendors like Micron, I've got very little incentive to go back to buying Samsung products.

    • cygnus1
    • 1 year ago

    I was a happy Prime Day customer and got the 1TB MX500 for only $155, or about 15.5 cents/GB. I’ll take that over the basically identical 860 EVO and it’s 22 cents/GB. The only better deal I’ve seen is on the Micro 1100 2TB drive down in the $250 range. Just don’t ‘need’ a 2TB boot disk at the moment

      • derFunkenstein
      • 1 year ago

      You do, however, need a 2TB Steam disk. 😉

        • cygnus1
        • 1 year ago

        Lol, that I won’t deny. My old 840 Evo currently fills that role and is 1TB. Not sure where the current 500GB boot SSD will go, but with the MX500 now I’ll have 2TB worth of SSD in the desktop at least 😀

          • oldog
          • 1 year ago

          I’ve now got a drawer full of “old” small capacity SSDs collecting dust. Join the club.

            • cygnus1
            • 1 year ago

            I’m starting to collect that too, especially my server SSDs. I’ve still got ancient hard drives sitting around on a closet shelf too, quite a few under 300GB

    • wierdo
    • 1 year ago

    SSD market feels kinda stagnant recently, performance is good enough territory and any more is just hitting deminishing returns big time.

    The only things left are price, size and reliability, probably in that order, and prices/sizes have not gone down as fast as one would hope. Reliability… no idea, everything is TLC and now they’re moving to QLC, so in theory it may have went down but hard to say if it matters for consumer market, might be good enough territory still.

    I have an old 960gb M500 and I don’t see anything on the market after all this time that would make me wanna upgrade yet.

    Come on guys, where’s the 4tb SSDs for same price as my old drive, it’s been many years, I thought that was the prediction back then, no?

      • strangerguy
      • 1 year ago

      It’s not just SSDs, but the whole consumer electronics market. The only thing that makes new stuff seem exciting anymore are low prices, instead of real ground-breaking innovation.

        • moose17145
        • 1 year ago

        Oh I legitimately like Micron/Intel’s 3D XPoint technology and all the things that Optane brings… but you are most correct in that price DOES matter immensly. As much as I like optane… it’s price precludes me from being willing to buy it (especially relative to it’s marginal performance benefits).

      • Waco
      • 1 year ago

      2 TB SSDs for $240 on sale is a pretty smoking deal IMO. I’ve certainly paid a lot more than that for (smaller) HDDs in the past.

        • chuckula
        • 1 year ago

        [url<]https://www.youtube.com/watch?v=wVOa3xhl0bg[/url<]

        • jihadjoe
        • 1 year ago

        > 2 TB SSDs for $240 on sale

        Where? Gimme gimme

          • Waco
          • 1 year ago

          Micron 1100 on eBay when they have a 20% off sale.

    • ermo
    • 1 year ago

    My old i7-3770k is rocking a 256GB 840 Pro as its boot drive.

    Just a few weeks ago, I picked up an 860 Evo for less than what that 840 Pro cost back in the day (the EVO was actually the cheapest 1TB SATA SSD for sale at my local hardware pusher).

    Can’t argue with the march of manufacturing progress.

    It’s going to be PC upgrade time soon, but I’m still waiting for DDR4 prices to drop a little lower.

      • jihadjoe
      • 1 year ago

      There is progress for sure, but it has to be noted that the NAND in your old 840 Pro was MLC (really ought to be called DLC now that quad-level NAND is in the works) and has much higher endurance than the TLC in your new 860 Evo.

        • strangerguy
        • 1 year ago

        IIRC Google themselves don’t care about SLC or MLC SSDs anymore as they found out that write endurance wasn’t a root cause for most of their SSDs that failed in production. Once more enterprise deployments follow suit SLC and MLC will be fading into obscurity.

          • jihadjoe
          • 1 year ago

          Of course they wouldn’t care. Google has redundancy up the wazoo.

          I think it was around 2009 they published a paper that says something about how data will be instantly replicated not just on multiple machines, but across multiple datacenters depending on use patterns.

          I bet there’s a kernel of truth in that [url=https://xkcd.com/1737/<]xkcd comic about massive redundancy[/url<].

    • strangerguy
    • 1 year ago

    So its a MX500 with worse GB/$ at MSRP and a lot worse in that metric than the 2TB Micron 1100 @ $250-300. Samsung needs to offer more than just their branding now IMO.

      • Waco
      • 1 year ago

      Yep. I’ve been rocking a Micron 1100 as my boot drive in the past few months – it’s “slower” than the one I replaced (SanDisk Extreme Pro) but I’ll be damned if there’s any difference in perceptible performance except that it’s 4 times as big for less money. 🙂

        • Freon
        • 1 year ago

        SSDs are completely commodity products now, just like HDDs. Performance is no longer a meaningful point of differentiation. Price is the only real factor.

        I think Samsung no longer holds an edge as others have figured out how to make fast, cheap TLC drives.

      • cygnus1
      • 1 year ago

      I got an MX500 1TB for $155 on Prime Day. I know that’s not the normal price, but it’s usually under $200. SATA drives are plateaued, throughput and IOPS aren’t getting any better on that type of connection, so it really comes down to GB/$ for SATA SSD’s. Samsung really just needs to drop the prices on the SATA models…

Pin It on Pinterest

Share This