Durability is the amount of data that can be written to the drive Samsung is guaranteeing can be written to it during its life. It's just related to the number of times a chip can be completely rewritten. If you double the capacity, you double the number of chips, and therefore you double the durability.
If i'm not mistaken that was a marketing move. The EVO was sold as 150 so that people who needed more would buy PRO. There was a super punishing SSD test half a year back (a bit less maybe) where they tested how much the drives can take before the cells give up and even the EVO held out beyond 300. So they didn't really manage to improve anything, they just decided to extend their guarantee.
EVO not as reliable as the Pro (TLC EVO vs MLC PRO,) the pro drive never failed with bad data it silently failed at 2PB
the endurance is for warranty before where they have found errors have cropped up, the PRO drive use MLC so is reliably less error prone (as in they handle any minor errors better then TLC) then EVO under extreme loads (drives might look the same but the flash is not)
sorry if this is been asked before, is the 25% OP been done via the ATA command that allows you to set amount of useable space or are you just partitioning it (both after a Secure erase)
i looked online i guess it worked as i thought it would, you can use ATA LBA limit command to hard limit the OP space or just make the partition say 20GB less then the size of the SSD (or what ever size you want) this is assuming its a fresh drive or secured erased state, but the thing is TRIM does the same thing any way (assuming you don't fill the drive so it runs out of space) if you cant use TRIM then you need to set higher OP space (recommended any way)
Not quite yet.... I've been limping along with a pair of 840 EVO's waiting for this Samsung 3D flash in an m.2 form factor with a PCIe x4 with NVMe controller. All of this technology is mature, and my motherboard has the m.2 slot and BIOS support for NVMe, but Samsung hasn't put it all together yet. I'm not shelling out any more money for SATA III SSDs when I feel that any time now...Samsung will introduce the product I've been hoping for. Am I foolish to think this product in imminent?
Hmm, I'm not sure anyone could really say... It's plausible they might skip retail availability of the SM951 AHCI & NVMe versions altogether and release something with 3D NAND this year, but I think they haven't released the former to begin with because the demand just isn't there outside of OEMs.
Might change after Skylake, might not, I doubt the average enthusiast is chomping at the bit to pay $1/GB for a drive but I dunno how much cheaper the SM951 would be in volume. I'm just aiming for a 256GB SM951 post-Skylake, for the OS, paired with 850 EVOs for storage.
with the 2TB drive they should be good up to 2-6PB of written data at least (the 150-300TB is for warranty purpose past that they will not cover it) the Pro drive would be recommended if you know you're going to be doing lots of writes as tends to keep on going till it fails with 0 errors (EVO does not bold as well under mass data loads when exceeding its endurance spec)
In the worst case at 14nm with the 840 Pro they saw ~800TB of writing before they acted up. They should easily see more than that with 40nm MLC 3d nand. 300TB is great way to to beat everyone else at the warranted writes and still only offer a warranty that covers 10% on the low of its projected use allowance. The 2 TB version is probably good for damn near 10 PB of writes (could be a lot more not using a calculator to figure out its true longevity).
since your average modern ssd will manage much more than 300tb anyways, upping the (guaranteed) durability from 150 to 300tb doesn't really take much and might as well be considered a marketing move. not that it's a bad development.
Even without the 3D V-NAND tech, when Samsung upped the warranty to five years on their 850 EVOs (compared to 3-years on the 840 EVOs) and still kept prices competitive, I was sold on their drives. If a company backs any of this PC tech for five years (and even 10 years on the 850 PRO series!) or more, their on the short list for any of my builds.
the length of their warranties sold me too. I've been recommending them to friends to. One of those friend's 840 Pro however has been having some problems, attempting to do an RMA, Samsung only said they'd send out a refurbished drive (and never an upgraded model). This has put quite a sour taste in my mouth with Samsung (sadly). I don't know if Intel or others are any better with their RMA SSDs, but I sure hope so.
It's common practice to send out refurbs for RMA replacement. For items under 30 days old, sometimes you can get a new replacement, depends on the manufacturer. What do you mean by upgraded model? Did he want an 850Pro for his 840 or something?
did a RMA recently and in the email there was a notice that said they give New units within 90 days of purchase as long as you provide the proof of purchase.
To be fair, SanDisk did the 10 years thing first on the Extreme Pro; Samsung may just be responding to that. The number of drives returned for warranty, even among drives that actually fail, is likely extremely low.
Hmm, already got a 1TB EVO... Chances are I'll find a better deal (GB/$) on a second one by the end of this year than on the 2TB. I wish these had been out for a while already, if it were $700 or less already I would've picked one up instead of the 1TB I got recently (for $340). Trying to ditch HDD for anything but backups, but it'll come down to how much storage my photography ends up gobbling thru the year.
The bug turned out to be in the Linux kernel, not in Samsung SSDs, as you can see in the first link once you scroll down the updates. Samsung has developed a kernel patch to fix the issue too.
IF you follow through to the mailing list discussion for the bug fix, the problem is with the kernel overwriting a pointer when it shouldn't be. If I'm following it correctly, it impacts any SSD brand in RAID0 with trim enabled.
There are two forms of TRIM these days. The original, Windows-supported, inline TRIM and the latest, queued TRIM. The latter is what is the problem on Samsung drives. I encourage you to fully investigate the issue.
Inline TRIM is known to cause delays with certain drives and certain host systems because it can take over IO on a drive and freeze other commands until TRIM is complete. The number of drives and systems effected is quite low, but it is enough for some people to disable TRIM or use a nightly TRIM script (fstrim).
That was my first reaction, too. But judging from the message on the mailing list and the patch, it is indeed a kernel issue and not specific to Samsung drives. It seems so stem from using queued TRIM on software RAID0, which is a moderately questionable configuration anyway. I guess Algolia did not tell the whole (probably embarrassing) story since there is only one mention of Linux software RAID in the entire article. Maybe they didn't configure their Intel drives the same way?
I was set on an Intel 730 for a 7mm SATA role up until a few minutes ago because I had read about this, too. But in light of this, one can probably use Kristian's "best 6Gbps SATA SSD" without excessive worry.
the bug is related to the incorrect Qued trim support on the Samsung drives
the samsung drives says they support Qued Trim support but they failed to implement it correctly when they added SATA 3.2 in the latest firmware updates, the Old firmware did not have Qued trim bug because the SSD did not advertise support for it, other SSDs that advertise Qued support it have been patched or don't have the buggy support to start off with (accept the m500)
i guess this is relating to RAID , there is a failed implementation of advertised Queued Trim support in the samsung 840 and 850 evo/pro drives (the drive says it supports it but it does not support it correctly so TRIM commands are issued incorrectly as to why there is a black list for all 840* and 850* drives)
your post seems to be related to RAID and kernel issue (but the issue did not happen on Intel drives that they changed to) they rebuilt there intel SSD setups the same as the samsung ones
they did the same auto restore only the drives changed they had 0 problems once they changed to intel/"other whatever it was" SSDs that also supported Qued TRIM even thought they was not using it the RAID bit probably was (was a bit ago when i looked at it)
the problem with samsung drives and Qued trim is till there (not just fake qued trim they failed to implement they also failed to implement the 3.2 spec and the advertised features that samsung is exposing)
Those two links show separate bugs. The algolia reported bug was a kernel issue. The second bug which vFunct was probably referring to is a firmware bug where the SSD advertises queued TRIM support but does not handle it correctly. The kernel works around this by blacklisting queued TRIM from known-bad drives. Windows doesn't support queued TRIM at all which is why you don't see the issue there yet.
Where did TRIM validation go? (The initial approach, which checked whether TRIM restored performance on a filled drive). Considering that controllers have had problems with TRIM not restoring performance, even if this is a minor revision, it still seems an important aspect to test.
The move to Windows 8 broke compatibility with that method since HD Tach no longer works. I do have another idea, though, but I just haven't had the time to try it out and implement it to our test suite.
It's mentioned at the end of 'The Destroyer' test that the 2TB 850 Pro uses less power than the 512GB variant, with one reason being cited as possibly a more efficient process node for the controller. Wouldn't it be more likely that the move to LPDDR3 in the 2TB variant was the cause for the increase in efficiency?
He said that on the final words page as well: "I'm very glad to see improved power efficiency in the 2TB models. A part of that is explained by the move from LPDDR2 to LPDDR3, but it's also possible that the MHX is manufactured using a more power efficient process node. "
Kristian, I like that you give a little bit of love to the Mixed Sequential Read/Write graphs.
Honestly this is the 1 area that I still find myself tearing out my hair waiting for on my Mid 2014 rMBP 1TB. I do a lot of work with large VMs in VMware and from time to time I have to copy one.
Peak read and Peak write speeds on this SSD are quite good, often approaching 1GB/s, but mixed sequential read/write is capped to an aggregate total of 1GB/s (yes I realize that this is bus limited on a x2 PCI-E 2.0 SSD).
This is one area that I really look forward to seeing improvements in with x4 PCI-E 3.0 SSDs.
Just wondered if you could shed some light onto the Mixed Sequential Read/Write significance for you. I'm not super-tecchy personally, so it doesn't mean very much to me in those words alone.
BUT, I have been experiencing some very frustrating behaviour on my rMBP late-2013 (with 1TB built in SS-storage). When I'm using Premiere Pro CC2015 with video projects over a certain size (I'm a pro cameraman and editor, so am using heavy XAVC-I video from a Sony FS-7), then I get crazy lags waiting for a sequence to open, or specifically when making copy&paste commands. I have noticed that my (16GB) RAM is often near full in these situations, and there is a swap file in action too (between 1-16GB).
Any thoughts on my problem? More specifically, any possible ideas/suggestions of a config adjustment that could improve my experience? Or, is it simply the case that I'm pushing my machine too hard, and need to get a 32GB-RAM-cabable laptop ASAP?
So they now have a way larger package/die for the pro version? Because with the 1TB evo and 1TB Pro i got the picture both are exactly the same hardware only the evo stored 3 in 1 cell and the pro 2
(128GBit / 3 * 2 = 86GBit pro)
But now they both have 128GBit for the evo this i guess just means more of the same stuff But for the Pro this has to mean that it has way more cells 50% more. So the die of the 2TB has to be 50% bigger then a one of the 1TB right?
The 86Gbit MLC and 128Gbit TLC dies are not identical -- the TLC die is actually smaller (68.9mm^2 vs 87.4mm^2) due to it being a single-plane design. A lot more than the number of memory transistors goes into the die size, so estimating the die size based on the 50% increase in memory capacity alone isn't really possible.
the max speed of SATA interface (any thing higher sequential 500mb/s sequential i would not even bother to look at sequential speed tests) with SSDs its the random speeds you should be looking at, as its not the sequential speed that makes a SSD fast (that is depending what your doing that is as a SSD sequential is still 3x faster then a HDD) as random access can be 20-80mb/s on SSD (random access tends to be as low as 0.5mb/s on a HDD or lower)
600MB/s is the theoretical maximum after 8b/10b encoding. However, there are other overheads (e.g. AHCI) that lead the maximum performance in real world being about 560MB/s.
Very good review, especially the wide gamut of the tests and the ability to see graphs for competing devices.
However, you still have left out: file copy: single stream and multiple stream.
The tests on: http://www.anandtech.com/show/9451/the-2tb-samsung... The 2TB Samsung 850 Pro & EVO SSD Review Mixed Random Read/Write Performance and http://www.anandtech.com/show/9451/the-2tb-samsung... The 2TB Samsung 850 Pro & EVO SSD Review SEQUENTIAL PERFORMANCE are good, but the reviews should include COPY tests. (Write Sequentially and Read Sequentially are not the same as COPY sequentially, and many devices have much less than 1/2 of the performance that would be expected if the devices were "perfect". For example if read sequential and write sequential were 550MB/s, one would expect at least 275MB/s for COPY, but might see only 100MB/s. [Since you will see that there some devices are able to copy at speeds above 250MB/s, you would know that drop in the copy speed to less than 1/2 of the write speed and read speed is a limit of the device under test, and not due to your configuration.] )
However Samsung 850 Pro 2TB 25% OP - 4KB Random Write (QD32) Performance used less than 37000 writes/second *4K bytes/write* 3400 seconds Rounding up, I calculate that the total data written, excluding the write amplification factor is less than 4E4 writes/second*4E3 seconds* 4KB/write = 1.6E8 *4KB so less than 160 M *4KB
The over-provisioning was at least 512GB= 512M *4KB So, unless the write amplification factor was larger than about 3 the steady state hasn't been reached. I expect that the write amplification factor was less than 2; given the amount of over provision it likely less than 1.25. Therefore I think at least 3 hour runtime is needed, maybe 4 hours. 6 hours seems like longer than needed.
Note that the Samsung 850 Pro 2TB likely to be able to erase blocks for use at a much higher speed than the external write speed that is being measured, but the still could be a change in performance at some point.
Nitpick: The packaging and labels on all my 850 PRO stuff (up to 1TB) say "PRO" not "Pro".
For enterprise drives, we run a longer six-hour 4KB random write test to measure steady-state to ensure that it's really steady state. For client drives an hour of 4KB random writes at QD32 is already unrealistic and it gives us data that is a fairly accurate representation of steady-state performance while keeping the test duration shorter to increase our test throughput.
Where are the Intel SSDs for comparison? Historically Intel drives have the most consistent drive performance across the entire drive cycle and use cases. However the Intel 730, and 500 series SATA drives are omitted from the comparisons. Why?
I've only had limited time to put drives through our new 2015 Client SSD Suite and given that Intel drives haven't been the most competitive in the SATA client space for a while, they haven't been my top priority.
>The Pro has been holding the performance crown for the past year and it's starting to look like no SATA drive can dethrone it
Probably. However, I'd continue to overlook the recommendations of Samsung products when there exist alternative products (such as from SanDisk) with comparable performance plus for the lesser $ price. I've no interest in paying more to companies like Samsung, thus making them more rich and larger.
Note: I own many many SSDs of various types and makes. In my R&D and Manufacturing business domain, IT product performance matters but there is a price thresold at which one shouldn't cross (RoI) while procurding those products in bulk. It just doesn't make any sense to ignore $ saving per product each year. Note that Samsung 850 crashed their prices only a little while ago.
The EVO has had a mild to substantial price advantage on everything else for a few months now tho... Kinda makes the Pro's premium moot unless you're in some sorta edge case where the difference between them matters and there's some equivalent alternative that happens to be priced just between the two.
As I see it, Samsung is pushing really hard to win over the whole market. The 850 Evo line is a substantial step forward. I really like the way PC starts working with an SSD rather than with HDD. I've bought one here http://hardware.nl/harde-schijven/samsung/ssd/mz-7... But you may find even a better offer on eBay, I think. Anyway, even for 85 euro this is a great purchase.
Can someone please explain the "Performance Consistenty" charts and how to read them? They change every time I switch the default and 25% OP even with the same drives... confusing.
I purchased the Samsung SSD 850 Pro to install in my new Windows 10 computer. I can't clone Windows 10 or anything to this drive. I've talked to Samsung for over 7 weeks and keep getting a new date for the release of their cloning software. I've tried 2 different cloning software and they don't work with this drive. Until Samsung updates the software for Windows 10 DON'T BUY.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
66 Comments
Back to Article
twizzlebizzle22 - Thursday, July 23, 2015 - link
Damn, we are in the future here. Roll in the day when my media storage drives are all SSD.Interesting as to why they managed to double the drive durability from 150TB to 300TB. That's pretty substantial.
joex4444 - Thursday, July 23, 2015 - link
Durability is the amount of data that can be written to the drive Samsung is guaranteeing can be written to it during its life. It's just related to the number of times a chip can be completely rewritten. If you double the capacity, you double the number of chips, and therefore you double the durability.lilmoe - Thursday, July 23, 2015 - link
Sure, but Samsung is also setting the older 512GB and 1TB at 300TB endurance rating up from 150TB previously, which is nice.SleepyFE - Friday, July 24, 2015 - link
If i'm not mistaken that was a marketing move. The EVO was sold as 150 so that people who needed more would buy PRO. There was a super punishing SSD test half a year back (a bit less maybe) where they tested how much the drives can take before the cells give up and even the EVO held out beyond 300. So they didn't really manage to improve anything, they just decided to extend their guarantee.leexgx - Saturday, July 25, 2015 - link
EVO not as reliable as the Pro (TLC EVO vs MLC PRO,) the pro drive never failed with bad data it silently failed at 2PBthe endurance is for warranty before where they have found errors have cropped up, the PRO drive use MLC so is reliably less error prone (as in they handle any minor errors better then TLC) then EVO under extreme loads (drives might look the same but the flash is not)
leexgx - Thursday, July 23, 2015 - link
sorry if this is been asked before, is the 25% OP been done via the ATA command that allows you to set amount of useable space or are you just partitioning it (both after a Secure erase)Samus - Thursday, July 23, 2015 - link
You don't even need to go that far. The Samsung Magician toolbox allows you to set a RAW OP "partition"leexgx - Thursday, July 23, 2015 - link
i looked online i guess it worked as i thought it would, you can use ATA LBA limit command to hard limit the OP space or just make the partition say 20GB less then the size of the SSD (or what ever size you want) this is assuming its a fresh drive or secured erased state, but the thing is TRIM does the same thing any way (assuming you don't fill the drive so it runs out of space) if you cant use TRIM then you need to set higher OP space (recommended any way)TEAMSWITCHER - Thursday, July 23, 2015 - link
Not quite yet.... I've been limping along with a pair of 840 EVO's waiting for this Samsung 3D flash in an m.2 form factor with a PCIe x4 with NVMe controller. All of this technology is mature, and my motherboard has the m.2 slot and BIOS support for NVMe, but Samsung hasn't put it all together yet. I'm not shelling out any more money for SATA III SSDs when I feel that any time now...Samsung will introduce the product I've been hoping for. Am I foolish to think this product in imminent?Impulses - Thursday, July 23, 2015 - link
Hmm, I'm not sure anyone could really say... It's plausible they might skip retail availability of the SM951 AHCI & NVMe versions altogether and release something with 3D NAND this year, but I think they haven't released the former to begin with because the demand just isn't there outside of OEMs.Might change after Skylake, might not, I doubt the average enthusiast is chomping at the bit to pay $1/GB for a drive but I dunno how much cheaper the SM951 would be in volume. I'm just aiming for a 256GB SM951 post-Skylake, for the OS, paired with 850 EVOs for storage.
leexgx - Thursday, July 23, 2015 - link
with the 2TB drive they should be good up to 2-6PB of written data at least (the 150-300TB is for warranty purpose past that they will not cover it) the Pro drive would be recommended if you know you're going to be doing lots of writes as tends to keep on going till it fails with 0 errors (EVO does not bold as well under mass data loads when exceeding its endurance spec)Topweasel - Friday, July 24, 2015 - link
In the worst case at 14nm with the 840 Pro they saw ~800TB of writing before they acted up. They should easily see more than that with 40nm MLC 3d nand. 300TB is great way to to beat everyone else at the warranted writes and still only offer a warranty that covers 10% on the low of its projected use allowance. The 2 TB version is probably good for damn near 10 PB of writes (could be a lot more not using a calculator to figure out its true longevity).fokka - Saturday, July 25, 2015 - link
since your average modern ssd will manage much more than 300tb anyways, upping the (guaranteed) durability from 150 to 300tb doesn't really take much and might as well be considered a marketing move. not that it's a bad development.Ubercake - Thursday, July 23, 2015 - link
Even without the 3D V-NAND tech, when Samsung upped the warranty to five years on their 850 EVOs (compared to 3-years on the 840 EVOs) and still kept prices competitive, I was sold on their drives. If a company backs any of this PC tech for five years (and even 10 years on the 850 PRO series!) or more, their on the short list for any of my builds.colinstu - Thursday, July 23, 2015 - link
the length of their warranties sold me too. I've been recommending them to friends to. One of those friend's 840 Pro however has been having some problems, attempting to do an RMA, Samsung only said they'd send out a refurbished drive (and never an upgraded model). This has put quite a sour taste in my mouth with Samsung (sadly). I don't know if Intel or others are any better with their RMA SSDs, but I sure hope so.hansmuff - Thursday, July 23, 2015 - link
It's common practice to send out refurbs for RMA replacement. For items under 30 days old, sometimes you can get a new replacement, depends on the manufacturer.What do you mean by upgraded model? Did he want an 850Pro for his 840 or something?
sustainednotburst - Friday, July 24, 2015 - link
did a RMA recently and in the email there was a notice that said they give New units within 90 days of purchase as long as you provide the proof of purchase.BillyONeal - Thursday, July 23, 2015 - link
To be fair, SanDisk did the 10 years thing first on the Extreme Pro; Samsung may just be responding to that. The number of drives returned for warranty, even among drives that actually fail, is likely extremely low.Impulses - Thursday, July 23, 2015 - link
Hmm, already got a 1TB EVO... Chances are I'll find a better deal (GB/$) on a second one by the end of this year than on the 2TB. I wish these had been out for a while already, if it were $700 or less already I would've picked one up instead of the 1TB I got recently (for $340). Trying to ditch HDD for anything but backups, but it'll come down to how much storage my photography ends up gobbling thru the year.melgross - Thursday, July 23, 2015 - link
The article is wrong about 1TB drives not being around at $1,000 and above. There were plenty of those drives.vFunct - Thursday, July 23, 2015 - link
Any info about the well known TRIM bug in these drives?vFunct - Thursday, July 23, 2015 - link
TRIM bug reported here: https://blog.algolia.com/when-solid-state-drives-a...and here:
https://bugs.launchpad.net/ubuntu/+source/fstrim/+...
Kristian Vättö - Thursday, July 23, 2015 - link
The bug turned out to be in the Linux kernel, not in Samsung SSDs, as you can see in the first link once you scroll down the updates. Samsung has developed a kernel patch to fix the issue too.BillyONeal - Thursday, July 23, 2015 - link
Well they patched the kernel to work around the firmware bug; but that doesn't mean it was a kernel bug.Kristian Vättö - Thursday, July 23, 2015 - link
There was never a problem with TRIM under Windows or OS X.DanNeely - Thursday, July 23, 2015 - link
IF you follow through to the mailing list discussion for the bug fix, the problem is with the kernel overwriting a pointer when it shouldn't be. If I'm following it correctly, it impacts any SSD brand in RAID0 with trim enabled.leexgx - Thursday, July 23, 2015 - link
did not affect the Intel SSDsmooninite - Thursday, July 23, 2015 - link
Kristian,There are two forms of TRIM these days. The original, Windows-supported, inline TRIM and the latest, queued TRIM. The latter is what is the problem on Samsung drives. I encourage you to fully investigate the issue.
Inline TRIM is known to cause delays with certain drives and certain host systems because it can take over IO on a drive and freeze other commands until TRIM is complete. The number of drives and systems effected is quite low, but it is enough for some people to disable TRIM or use a nightly TRIM script (fstrim).
sustainednotburst - Friday, July 24, 2015 - link
Algolia stated Queued Trim is disabled on their systems, so its not related to Queued Trim.editorsorgtfo - Thursday, July 23, 2015 - link
That was my first reaction, too. But judging from the message on the mailing list and the patch, it is indeed a kernel issue and not specific to Samsung drives. It seems so stem from using queued TRIM on software RAID0, which is a moderately questionable configuration anyway. I guess Algolia did not tell the whole (probably embarrassing) story since there is only one mention of Linux software RAID in the entire article. Maybe they didn't configure their Intel drives the same way?I was set on an Intel 730 for a 7mm SATA role up until a few minutes ago because I had read about this, too. But in light of this, one can probably use Kristian's "best 6Gbps SATA SSD" without excessive worry.
leexgx - Thursday, July 23, 2015 - link
the bug is related to the incorrect Qued trim support on the Samsung drivesthe samsung drives says they support Qued Trim support but they failed to implement it correctly when they added SATA 3.2 in the latest firmware updates, the Old firmware did not have Qued trim bug because the SSD did not advertise support for it, other SSDs that advertise Qued support it have been patched or don't have the buggy support to start off with (accept the m500)
editorsorgtfo - Thursday, July 23, 2015 - link
Can you corroborate this? Nothing in the patch hints at a vendor issue.leexgx - Thursday, July 23, 2015 - link
i guess this is relating to RAID , there is a failed implementation of advertised Queued Trim support in the samsung 840 and 850 evo/pro drives (the drive says it supports it but it does not support it correctly so TRIM commands are issued incorrectly as to why there is a black list for all 840* and 850* drives)your post seems to be related to RAID and kernel issue (but the issue did not happen on Intel drives that they changed to) they rebuilt there intel SSD setups the same as the samsung ones
they did the same auto restore only the drives changed they had 0 problems once they changed to intel/"other whatever it was" SSDs that also supported Qued TRIM even thought they was not using it the RAID bit probably was (was a bit ago when i looked at it)
sustainednotburst - Friday, July 24, 2015 - link
Algolia stated Queued Trim is disabled on their systems, so its not related to Queued Trim.leexgx - Saturday, July 25, 2015 - link
the problem with samsung drives and Qued trim is till there (not just fake qued trim they failed to implement they also failed to implement the 3.2 spec and the advertised features that samsung is exposing)Gigaplex - Thursday, July 23, 2015 - link
Those two links show separate bugs. The algolia reported bug was a kernel issue. The second bug which vFunct was probably referring to is a firmware bug where the SSD advertises queued TRIM support but does not handle it correctly. The kernel works around this by blacklisting queued TRIM from known-bad drives. Windows doesn't support queued TRIM at all which is why you don't see the issue there yet.jann5s - Thursday, July 23, 2015 - link
@AT: please do some data retention measurements with SSD drives! I'm so curious to see if the myth is true and to what extent!Shadowmaster625 - Thursday, July 23, 2015 - link
With 2 whole gigabytes of DRAM, why are random writes not saturating the SATA bus?Kristian Vättö - Thursday, July 23, 2015 - link
The extra DRAM is needed for the NAND mapping table, it's not used to cache any more host IOs.KaarlisK - Thursday, July 23, 2015 - link
Where did TRIM validation go? (The initial approach, which checked whether TRIM restored performance on a filled drive).Considering that controllers have had problems with TRIM not restoring performance, even if this is a minor revision, it still seems an important aspect to test.
Kristian Vättö - Thursday, July 23, 2015 - link
The move to Windows 8 broke compatibility with that method since HD Tach no longer works. I do have another idea, though, but I just haven't had the time to try it out and implement it to our test suite.editorsorgtfo - Thursday, July 23, 2015 - link
Kristian, what would you consider the best SATA 6Gbps drive(s) with power-loss protection?editorsorgtfo - Thursday, July 23, 2015 - link
Judging from the 850 Pro and EVO PCBs, they don't even guard their NAND mappings. Or my eyesight is giving.Meegul - Thursday, July 23, 2015 - link
It's mentioned at the end of 'The Destroyer' test that the 2TB 850 Pro uses less power than the 512GB variant, with one reason being cited as possibly a more efficient process node for the controller. Wouldn't it be more likely that the move to LPDDR3 in the 2TB variant was the cause for the increase in efficiency?MikhailT - Thursday, July 23, 2015 - link
He said that on the final words page as well: "I'm very glad to see improved power efficiency in the 2TB models. A part of that is explained by the move from LPDDR2 to LPDDR3, but it's also possible that the MHX is manufactured using a more power efficient process node. "MrCommunistGen - Thursday, July 23, 2015 - link
Kristian, I like that you give a little bit of love to the Mixed Sequential Read/Write graphs.Honestly this is the 1 area that I still find myself tearing out my hair waiting for on my Mid 2014 rMBP 1TB. I do a lot of work with large VMs in VMware and from time to time I have to copy one.
Peak read and Peak write speeds on this SSD are quite good, often approaching 1GB/s, but mixed sequential read/write is capped to an aggregate total of 1GB/s (yes I realize that this is bus limited on a x2 PCI-E 2.0 SSD).
This is one area that I really look forward to seeing improvements in with x4 PCI-E 3.0 SSDs.
jas.brooks - Sunday, March 13, 2016 - link
Hey MrCommunistGen,Just wondered if you could shed some light onto the Mixed Sequential Read/Write significance for you. I'm not super-tecchy personally, so it doesn't mean very much to me in those words alone.
BUT, I have been experiencing some very frustrating behaviour on my rMBP late-2013 (with 1TB built in SS-storage). When I'm using Premiere Pro CC2015 with video projects over a certain size (I'm a pro cameraman and editor, so am using heavy XAVC-I video from a Sony FS-7), then I get crazy lags waiting for a sequence to open, or specifically when making copy&paste commands. I have noticed that my (16GB) RAM is often near full in these situations, and there is a swap file in action too (between 1-16GB).
Any thoughts on my problem? More specifically, any possible ideas/suggestions of a config adjustment that could improve my experience? Or, is it simply the case that I'm pushing my machine too hard, and need to get a 32GB-RAM-cabable laptop ASAP?
Thanks!
jason
jcompagner - Thursday, July 23, 2015 - link
So they now have a way larger package/die for the pro version?Because with the 1TB evo and 1TB Pro i got the picture
both are exactly the same hardware only the evo stored 3 in 1 cell and the pro 2
(128GBit / 3 * 2 = 86GBit pro)
But now they both have 128GBit for the evo this i guess just means more of the same stuff
But for the Pro this has to mean that it has way more cells 50% more. So the die of the 2TB has to be 50% bigger then a one of the 1TB right?
Kristian Vättö - Saturday, July 25, 2015 - link
The 86Gbit MLC and 128Gbit TLC dies are not identical -- the TLC die is actually smaller (68.9mm^2 vs 87.4mm^2) due to it being a single-plane design. A lot more than the number of memory transistors goes into the die size, so estimating the die size based on the 50% increase in memory capacity alone isn't really possible.karakarga - Friday, July 24, 2015 - link
Why the read and write speeds not increasing? 550~540 Mb/s read and 520 MB/s write speeds, may reach 600 MB's there are still headroom!leexgx - Saturday, July 25, 2015 - link
the max speed of SATA interface (any thing higher sequential 500mb/s sequential i would not even bother to look at sequential speed tests)with SSDs its the random speeds you should be looking at, as its not the sequential speed that makes a SSD fast (that is depending what your doing that is as a SSD sequential is still 3x faster then a HDD) as random access can be 20-80mb/s on SSD (random access tends to be as low as 0.5mb/s on a HDD or lower)
Kristian Vättö - Sunday, July 26, 2015 - link
600MB/s is the theoretical maximum after 8b/10b encoding. However, there are other overheads (e.g. AHCI) that lead the maximum performance in real world being about 560MB/s.DIYEyal - Friday, July 24, 2015 - link
Great review as always!By the way, I found a typo in page 4: "or full details of the test, please refer to the this article."
mark53916 - Friday, July 24, 2015 - link
Very good review, especially the wide gamut of the tests and theability to see graphs for competing devices.
However, you still have left out:
file copy: single stream and multiple stream.
The tests on:
http://www.anandtech.com/show/9451/the-2tb-samsung...
The 2TB Samsung 850 Pro & EVO SSD Review
Mixed Random Read/Write Performance
and
http://www.anandtech.com/show/9451/the-2tb-samsung...
The 2TB Samsung 850 Pro & EVO SSD Review
SEQUENTIAL PERFORMANCE
are good, but the reviews should include COPY tests. (Write Sequentially
and Read Sequentially are not the same as COPY sequentially, and many devices
have much less than 1/2 of the performance that would be expected if the devices
were "perfect". For example if read sequential and write sequential were 550MB/s,
one would expect at least 275MB/s for COPY, but might see only 100MB/s.
[Since you will see that there some devices are able to copy at speeds above 250MB/s,
you would know that drop in the copy speed to less than 1/2 of the write speed
and read speed is a limit of the device under test, and not due to your
configuration.] )
Thank you, especially for the power graphs and over-provisioned performance data:
http://www.anandtech.com/show/9451/the-2tb-samsung...
The 2TB Samsung 850 Pro & EVO SSD Review
Performance Consistency
However
Samsung 850 Pro 2TB 25% OP - 4KB Random Write (QD32) Performance
used less than 37000 writes/second *4K bytes/write* 3400 seconds
Rounding up, I calculate that the total data written,
excluding the write amplification factor is less than
4E4 writes/second*4E3 seconds* 4KB/write = 1.6E8 *4KB
so less than 160 M *4KB
The over-provisioning was at least 512GB=
512M *4KB
So, unless the write amplification factor was larger than about
3 the steady state hasn't been reached. I expect that the write
amplification factor was less than 2; given the amount
of over provision it likely less than 1.25. Therefore I
think at least 3 hour runtime is needed, maybe 4 hours. 6
hours seems like longer than needed.
Note that the Samsung 850 Pro 2TB likely to be able to erase
blocks for use at a much higher speed than the external write
speed that is being measured, but the still could be a change
in performance at some point.
Nitpick: The packaging and labels on all my 850 PRO stuff (up to 1TB) say
"PRO" not "Pro".
Kristian Vättö - Sunday, July 26, 2015 - link
For enterprise drives, we run a longer six-hour 4KB random write test to measure steady-state to ensure that it's really steady state. For client drives an hour of 4KB random writes at QD32 is already unrealistic and it gives us data that is a fairly accurate representation of steady-state performance while keeping the test duration shorter to increase our test throughput.Navier - Sunday, July 26, 2015 - link
Where are the Intel SSDs for comparison? Historically Intel drives have the most consistent drive performance across the entire drive cycle and use cases. However the Intel 730, and 500 series SATA drives are omitted from the comparisons. Why?Kristian Vättö - Sunday, July 26, 2015 - link
I've only had limited time to put drives through our new 2015 Client SSD Suite and given that Intel drives haven't been the most competitive in the SATA client space for a while, they haven't been my top priority.akula2 - Monday, July 27, 2015 - link
>The Pro has been holding the performance crown for the past year and it's starting to look like no SATA drive can dethrone itProbably. However, I'd continue to overlook the recommendations of Samsung products when there exist alternative products (such as from SanDisk) with comparable performance plus for the lesser $ price. I've no interest in paying more to companies like Samsung, thus making them more rich and larger.
Note: I own many many SSDs of various types and makes. In my R&D and Manufacturing business domain, IT product performance matters but there is a price thresold at which one shouldn't cross (RoI) while procurding those products in bulk. It just doesn't make any sense to ignore $ saving per product each year. Note that Samsung 850 crashed their prices only a little while ago.
Impulses - Tuesday, July 28, 2015 - link
The EVO has had a mild to substantial price advantage on everything else for a few months now tho... Kinda makes the Pro's premium moot unless you're in some sorta edge case where the difference between them matters and there's some equivalent alternative that happens to be priced just between the two.TelstarTOS - Monday, July 27, 2015 - link
The 2TB EVO is a very interesting drive, perfect for a notebook.SviatA - Friday, August 21, 2015 - link
As I see it, Samsung is pushing really hard to win over the whole market. The 850 Evo line is a substantial step forward. I really like the way PC starts working with an SSD rather than with HDD.I've bought one here http://hardware.nl/harde-schijven/samsung/ssd/mz-7...
But you may find even a better offer on eBay, I think. Anyway, even for 85 euro this is a great purchase.
htwingnut - Saturday, September 19, 2015 - link
Can someone please explain the "Performance Consistenty" charts and how to read them? They change every time I switch the default and 25% OP even with the same drives... confusing.[email protected] - Thursday, October 15, 2015 - link
I purchased the Samsung SSD 850 Pro to install in my new Windows 10 computer. I can't clone Windows 10 or anything to this drive. I've talked to Samsung for over 7 weeks and keep getting a new date for the release of their cloning software. I've tried 2 different cloning software and they don't work with this drive. Until Samsung updates the software for Windows 10 DON'T BUY.Miller1331 - Tuesday, December 1, 2015 - link
Using 2 of these in a music production rig and they are monstersMeteor2 - Wednesday, November 23, 2016 - link
I don't think this reviews specifies what interface the 850 Pro uses anywhere.Meteor2 - Wednesday, November 23, 2016 - link
...until the last page. Would've thought it page 1 material, myself.