The SSD Review The Worlds Dedicated SSD Education and Review Resource | 
Some owners of Samsung 840 EVO SSDs were among an unfortunate group that began to experience performance degradation over time. (We previously covered this here and here) In a nutshell, their drives could not read “cold” data (that had sat unmodified for an extended period of time) at full speed. This issue thus far has been mostly limited to 840 EVO SSDs with 19nm TLC NAND (planar, not 3D-VNAND).
Image source: pcper.com
The first attempt at fixing the issue was via a combination of a firmware update and a software tool. The software tool initiated a one-time “refresh”, causing all stale data to be rewritten. Although this restored performance back to optimal speeds, many users were skeptical (and rightly so) that their drives would eventually deteriorate back to the same state over time.
Image source: pcper.com
Allyn Malventano over at PC Perspective has been involved in this issue from its first appearance, and has worked directly with Samsung in helping identify the underlying problem, and beta-testing the “fixes”. You can read his review of the combination of the latest firmware update, and of Samsung Magician 4.6 here. There is a nice little “Q & A” between Allyn and Samsung about what the firmware update and the new Samsung Magician 4.6 software are able to accomplish.
The initial conclusion is that the firmware update does its own periodic refreshing of data, to eliminate the “stale data” issue altogether. Along with the revised firmware there is an “Advanced Performance Optimization” function that can be triggered within Samsung SSD Magician. For owners who never tried the one-time “Performance Restoration Tool” stop-gap fix, they can skip that completely and update to the new firmware, and hopefully be good to go.
Image source: pcper.com
Once users have downloaded the latest Samsung Magician 4.6 software, they will see an “Advanced” tab under “Performance Optimization.” The “standard” optimization merely triggers the TRIM function to execute on all empty portions of the SSD. The “Advanced” performance optimization triggers the same action that was triggered by the previous “Performance Restoration Tool.” The original was for one-time usage; but this version available through Samsung SSD Magician can be triggered whenever, or as frequently as desired.
Image source: pcper.com
Allyn’s review demonstrates “before firmware update”, “after firmware update”, and “after firmware update and after advanced optimization” results. These results show that merely updating the firmware enables a large improvement in read speeds of the stale data. This may indicate that the read algorithm has been tweaked to compensate for “cell drift over time”.
Image source: pcper.com
Running the advanced optimization function refreshes the data (likely in a sequential pattern), showing that randomly written files are restored to an even better state than they were previously.
Image source: pcper.com
It took Samsung a little while, but it appears that they have finally resolved the 840 EVO performance degradation issue. In the process, Samsung has given the 840 EVO an added feature via the “Advanced Performance Optimization” tool. As Allyn puts it, “….so in trying to fix one problem, they have actually added a useful feature to this product line.” It will be interesting to see if this feature is made available for any other Samsung SSDs, particularly the “non-840 EVO” other SSDs that utilize TLC NAND.
We will provide links to the 840 EVO firmware update, and to the Samsung Magician 4.6 software download once they are released into the wild.
Well done, Allyn!
UPDATE: Please visit the latest article on this topic for links (at end of article) for the latest firmware update and the latest Magician 4.6 software:
>Performance Degradation Fixed At Last?
Nope, they just slapped a bandaid on the problem. You can’t fix physics with firmware.
while I agree with you, the final output really is fixed performance problem (although not exactly “the optimal way” but what can you do with millions of products in the wild).
It would be great, simply just great, if Samsung does NOT include writes introduced with this refresh algorithm into overall Total Bytes Written counter – using another words “this refresh algorithm should not decrease the useful warranty for customer”. To be honest, 99.99999% of all customers in the world will never come close to physical endurance of these drives and will not wear the drives out, so this doesn’t represent any real affair to be afraid of. No conspiracies here necessary.
Well those three or four people that do 🙂 should have not bought EVOs in the first place. Any professional heavy-writer definitely should not rely on TLC technology at all, and if he does, he should bear consequences of buying something 30$ cheaper his business, money and productivity relies on. That is the difference between EVOs and PROs with the same capacity. And I really, really don’t want to get started about PRODUCTIVITY caused by better performance of PRO models [compared to EVOs] so even this higher CapEx would be far more than outweighed by productivity increase.
Cheers !
Does samsung even limit warranty by TBW ?
Benjamin, I don’t have direct experience, but I guess every manufacturer states TBW for some reason. It is the easiest excuse to refuse warranty on one side, on the other side – as we already said – just extremely tiny portion or amount of customers will ever hit that wall.
From 850Pro review on some anonymous site:
“The other major improvement from V-NAND is the endurance. All
capacities, including the smallest 128GB, are rated at 150TB, which is
noticeably higher than what any other consumer-grade SSD offers.
Moreover, Samsung told me that the endurance figure is mainly meant to
separate the 850 Pro from the enterprise drives to guide enterprise
clients to the more appropriate (and expensive) drives as the 850 Pro
does not have power loss protection or end-to-end data protection for
example. However, I was told that the warranty is not automatically
denied if 150TB is reached under a client workload. In fact, Samsung
said that they have a 128GB 850 Pro in their internal testing with over
eight petabytes (that is 8,000TB) of writes and the drive still keeps
going, so I tip my hat to the person who is able to wear out an 850 Pro
in a client environment during my lifetime.”
they may say warranty is not automatically denied if more than TBW is actually written (and it doesn’t matter who says that or how much TBW is), but who knows what really happens when rubber hits the road. In case of Sammies 850Pro and several other models, many capacities share the same TBW which – from engineering point of view – is just nonsense. Bigger capacities must have more endurance.
In the end, I’d say it’s a matter of luck : someone had good day or bad one when he hits your RMA SSD with exhausted TBW.
Enjoy !
It has 3 years warranty, so if my 840 EVO dies before that it should not matter even if it reached 1 PB.
From Where can we get this Magician 4.6?
As the end of this article states, it’s not released yet. You will be able to get it from sites like this, or from Samsung.com. You might even get notified by your existing copy of Samsung Magician that an update is available, which is the normal procedure for new versions of Magician.
It seems that TLC NAND isn’t the only type of NAND that’s having a problem at process nodes of 19nm of less. The Crucial MX200 SSD which uses 16nm MLC 128Gb NAND flash chips appears to be having the same very issue that the Samsung 840 EVO had, namely NAND flash voltage drifting which causes read speed performance issues on older data.
What does this mean? Well, it means that pretty much NAND flash memory becomes unreliable as the process node shrinks past 20nm. NAND flash manufacturers have hitthe wall when it comes to planar NAND. From this point on the only way to make NAND flash reliable is to reverse the process node shrinkage and go to 3D-NAND where NANDcells are stacked on one another instead of laid out flat.
Samsung started this trend in which they released their 3D-VNAND and reversed the process shrinkage to 40nm with their 850 line of SSDs. Other manufacturers are also doing the same with their own approach to 3D-NAND. Ultimately it comes down to the fact that NAND flash memory manufacturers have hit the wall in terms of shrinking NAND and that as you shrink the process node the NAND has far more of a chance of voltage drifting that can effect read speed performance as the data ages.