SSD Endurance Questions

Topic

New Reply

Do virus scans shorten the life of an SSD drive? I like to regularly scan with a couple of tools in addition to Windows Defender which runs by default. I am not sure if these use up valuable writes to the SSD, thus wearing it out?

Secondly, how can you tell how many writes you have left or what percentage? For my relatively new drive, CrystalDiskInfo has “Total Host Writes” of 670GB so assume this is the figure I need to keep an eye on. My drive (Crucial MX500 500GB) is only a couple of weeks old, and I notice that on the Crucial website the drive has a SSD Endurance (TBW) of 360 terabytes. So if I am at 670GB I assume I am not even up to 1TB yet? I notice that the 2TB MX500 drive has the same SSD Endurance (TBW) of 360 terabytes. Is it normal that different capacities have the same TBW? I thought the larger the drive would be able to have more data written to it?

And I assume there is no reason why the drive should not outperform the 360TBW figure? Hopefully significantly more, or perhaps less. You just have to backup I suspect and wait to see.

Thanks

Reply | Quote

Replies

So if you completely rewrote your 670Gb every day your drive would last 240 days. Assuming that you only rewrite about 20% of that every day (134Gb) you drive will last 240×5 or 1,200 days that’s 3.28 years. Now that’s a lot of writing and on a very consistent basis so from a reality standpoint you should get at least 5 years of service from the drive.

Generally, the life span concerns of SSD failing are overblown. That said, when SSDs go they GO! No warning like with spinners. Moral BACKUP, BACKUP, BACKUP!

HTH

May the Forces of good computing be with you!

RG

PowerShell & VBA Rule!
Computer Specs

2 users thanked author for this post.

teuhasn2, windbg

Reply | Quote

RetiredGeek wrote:

Generally, the life span concerns of SSD failing are overblown. That said, when SSDs go they GO! No warning like with spinners.

SSDs all have some sort of counter you can see counting down from day 1, sort of like a gas gauge, which you cannot see in a hard drive. Reaching that limit does not mean that the drive is “done,” though, unless its manufacturer has put a time bomb into the firmware to turn the rating into a hard limit (as Intel has done with at least some of their drives).

In terms of media wearout, you can watch as the bad blocks in a SSD accumulate in the SMART data. As each of them is removed from service, it will be replaced with one from the reserve, and the drive will keep working. Bad blocks accumulating smoothly isn’t a cause for alarm in SSDs, but it is an indicator of the media state in the drive. In theory, the drive should be able to keep going until it has exhausted all of the reserve blocks, but that would be only if the media wearout was the only failure mode, and it isn’t.

An SSD can also fail if the drive electronics fail (not counting the NAND chips). In this case, it will likely be an instant “lights out” kind of death. Hard drives have electronics too, though, and if there is a failure in the “logic board” as it is typically called, you will probably see the instant death just as on a SSD.

The most recent HDD failure I experienced was that type… I was installing Windows 7 on my Asus laptop’s nearly 5 year old Seagate HDD, and partway through, it just insta-bricked. No warning, no noises, no signs at all in the SMART data (which I had checked before I began the installation).

One should not assume there will be some kind of warning before a drive dies (and takes your data with it), even if it is a hard drive. Stay alert for the signs of trouble, but remember that sudden unexplained drive death does happen, and so do power surges (lightning, perhaps) and power supply failures, not to mention malware, bad updates, bugs in disk utilities, or accidents on the part of the user.

I was tasked with picking up the pieces in the early to mid 90s when one of the company’s AST computers failed. There wasn’t much left alive inside that PC… the CPU was dead, the motherboard dead, the hard drive dead, the PSU dead. The RAM and the floppy drive still worked, though I am not sure how much I would trust that RAM either, given how much else had failed on the PC.

In the bottom of the case were a number of small chunks of black plastic. I soon found that various chips on the motherboard had literally exploded, blowing the plastic bits off. My guess is that the PSU failed catastrophically and massively overvolted the system, perhaps even sending 110v AC into one or more of the power rails. There was no report of an electrical storm associated… it just blew up (figuratively; there was no damage outside the case) one day.

 

Dell XPS 13/9310, i5-1135G7/16GB, KDE Neon 6.2
XPG Xenia 15, i7-9750H/32GB & GTX1660ti, Kubuntu 24.04
Acer Swift Go 14, i5-1335U/16GB, Kubuntu 24.04 (and Win 11)

1 user thanked author for this post.

windbg

Reply | Quote

Do virus scans shorten the life of an SSD drive?

Since a scan does not involve writing to the drive, it will not shorten the life of the drive.

 

Most SSDs have 700 TB writes before it turns to read only, but the larger drives shorten this number of writes to about half (or as you found out – 360 TB).  This is the price for the speed of the SSDs.  That being said, it would take you years to have that many writes to the drive.

You are correct with CrystalDiskInfo  Total Host Writes.  you have used 670 GB of your 360 TB.  Since it takes 1000 GBs to = 1 TB, you have 359.33 TB writes left or 99.8 % of life left.

If you want to increase the life of an SSD, install a second platter (SATA) drive to keep your data on and keep only applications on the SSD.  Most writes come from adding and modifying personal data.

 

HTH, Dana:))

HTH, Dana:))

1 user thanked author for this post.

windbg

Reply | Quote

The main advantage of an SSD is write speed. Put all your programs and data on your SSD to maximize this benefit.

Relax in the knowledge that you have a really fast machine and it is protected by your backups.

cheers, Paul

Reply | Quote

anonymous wrote:

Do virus scans shorten the life of an SSD drive?

No. You can read all you want and not shorten the drive’s life, and scanning is reading.

anonymous wrote:

Secondly, how can you tell how many writes you have left or what percentage? For my relatively new drive, CrystalDiskInfo has “Total Host Writes” of 670GB so assume this is the figure I need to keep an eye on. My drive (Crucial MX500 500GB) is only a couple of weeks old, and I notice that on the Crucial website the drive has a SSD Endurance (TBW) of 360 terabytes. So if I am at 670GB I assume I am not even up to 1TB yet?

You are correct… you haven’t gotten to 1TB yet.

As far as percentages, there’s no standard way of reporting this. Each drive manufacturer has its own way of doing things. My Samsung drives use attribute 177, Wear Leveling Count, as the percentage-based remaining life of each drive.

anonymous wrote:

I notice that the 2TB MX500 drive has the same SSD Endurance (TBW) of 360 terabytes. Is it normal that different capacities have the same TBW? I thought the larger the drive would be able to have more data written to it?

You are right in thinking a 2TB drive would have a longer expected life than the 500GB (four times the TBW), but note that these figures from the manufacturer are just ratings, not predictions of life expectancy. SSD manufacturers typically apply a warranty that goes for 2-5 years or a certain number of TBW, and that warranty length may be the actual reason behind the lower write endurance per unit of storage that the 2TB drive has.

anonymous wrote:

And I assume there is no reason why the drive should not outperform the 360TBW figure?

I would certainly expect it to! TechReport tested a number of SSDs by writing to them continuously until they failed, and the Samsung 840 Pro (which I remember because it was the “winner” and because it was the same model of drive that I have in my desktop PC) went to about 2 PB before it failed, waaaaay above its rated service life.

Some Intel drives might be the exception, as the one in the TR test had sort of a time bomb in the firmware that would detect when the drive has reached its rated life and brick the drive (not just make it read-only… it makes it read-only instantly when it reaches the limit, but the next power cycle kills it, so you better be on your toes and ready to rescue everything the moment the drive stops accepting writes!). That’s… awful. I do not know if current Intel drives still have that “feature.”

 

 

Dell XPS 13/9310, i5-1135G7/16GB, KDE Neon 6.2
XPG Xenia 15, i7-9750H/32GB & GTX1660ti, Kubuntu 24.04
Acer Swift Go 14, i5-1335U/16GB, Kubuntu 24.04 (and Win 11)

Reply | Quote

Another No vote.  We have, let’s see, ten SSD’s, the oldest going back to 2015, newest 2020 and they all have over 98% life left, most 100%  Lots of gaming, videos, daily on time.  Half NVMe half SATA.  Kinda wish the urban legends going back to 1492 would die.

The only computer we have with mechanical drives is our home server, a Dell T30 with two Seagate Ironwolf 6 tb drives.  They’d be SSD’s if drives that size were available for less than the price of a car when I put it together.

The drive you have is a good one, try TechPowerUp or usrbenchmark for some decent info.

https://www.techpowerup.com/

https://hdd.userbenchmark.com/

HWinfo can give a good life estimate by reading the data the drive outputs without running stress tests and info about all the other components.  Portable version is fine.

https://www.hwinfo.com

Reply | Quote

Another No vote.  Of our ten SSD, some going back to 2015, all are at 98%  or more, most at 100% life.  Try hwinfo for a good utility to track life and all the other components. Portable.

Not posting any more links today, they get obliterated or my comment flies into the ether.

Reply | Quote

Your post made it into the spam bucket because you edited too fast / many times. Hold back on the edits and all will be well.

cheers, Paul

1 user thanked author for this post.

SallyBrown

Reply | Quote

Thanks, thought something like that.  Kept inserting the link for HWinfo in a strange format.

Reply | Quote

SSD Reliability: Is Your SSD Less Reliable Than A Hard Drive?

Reliability and Understanding SSD Lifespan
“Do SSD drives fail?” Well, that’s actually redundant since the D stands for drives, but yes. Of course. Everything fails. Rather we should be asking: “Do SSD drives fail faster?” That is a bit more complicated though.

SSDs are more reliable when it comes to harsh environments than HDDs because they don’t have actuator arms or any moving parts. As such, SSDs can withstand accidental drops and extreme temperatures better than HDDs…

Let’s look at the Samsung SSD 850 Pro SATA with a 10 year warranty that has a capacity of 128GB, 256GB, 512GB, and up to 1TB. This is designed to handle 150 terabytes written, or 150 TBW. But according to Samsung, these SSDs can withstand up to 600 TBW.

Now, if we’re going to do the math and calculate the endurance of this SSD over the course of 10 years, this drive can read and write 40 GB of data a day. Reflecting this to a real-life situation, where a normal office user writes between 10 GB to 35 GB a day, the warranty would have expired by the time you reach the 150 TBW limit…

MTBF, or Mean Time Between Failures
SSD vs HDD reliability rate is defined by MTBF, or Mean Time Between Failures. This is one of the more tricky, and less obvious reliability measurements. It’s important that you understand MTBF and solid state drive failure rates before making any type of purchases.

Let’s take an Intel® SSD 335 Series (240GB, 2.5in SATA 6Gb/s, 20nm, MLC) and SAMSUNG 830 Series 2.5″ 256GB SATA III MLC. The Intel SSD has an MTBF spec of 1.2 million hours while the Samsung SSD has an MTBF spec of 1.5 million hours. But this doesn’t mean that because Samsung has more hours that it will last longer– MTBF doesn’t work that way…

Reply | Quote

One of the most concise and clear explanations about SSD life expectancy is from Christopher Barnatt and his YouTube channel ExplainingComputers. The video is titled SSD Life Expectancy. The video is four years old but it addresses most of the technologies that remain in use today for SSDs and is valid in every respect. Barnatt made a slightly updated video in 2020 titled Explaining SSDs: Form Factors, Interfaces & Technologies.

A key takeaway: The larger the drive, the greater the amount of over-provisioning and the longer the life of the drive. Therefore, buy the largest SSD you can afford even if you don’t need that much space.

2 users thanked author for this post.

Sueska, OscarCP

Reply | Quote

Great video, with a clear, and to this non-SSD expert, understandable explanation virtually free of insiders’ talk or comments obscure to non-insiders.

But I still would like to know what SallyBrown, who seems to know what this is all about, meant here in this thread ( #2382112 ) by their 10 SSDs, some dating from 2015 (such as the one in my Mac) being at 80% and 100%.

All the information I have on how my Mac’s SSD is doing (without first installing some mystery application, or performing some elaborate ritual) is from “About this Mac/System Report/Storage”, where the relevant part is where it reads that the SSD has been “Verified”, a good thing to know, the alternative being “Failing” and “Failed” (one that I may never get to see, for obvious reasons.)

Ex-Windows user (Win. 98, XP, 7); since mid-2017 using also macOS. Presently on Monterey 12.15 & sometimes running also Linux (Mint).

MacBook Pro circa mid-2015, 15" display, with 16GB 1600 GHz DDR3 RAM, 1 TB SSD, a Haswell architecture Intel CPU with 4 Cores and 8 Threads model i7-4870HQ @ 2.50GHz.
Intel Iris Pro GPU with Built-in Bus, VRAM 1.5 GB, Display 2880 x 1800 Retina, 24-Bit color.
macOS Monterey; browsers: Waterfox "Current", Vivaldi and (now and then) Chrome; security apps. Intego AV

Reply | Quote

Looking at the data HWinfo puts out for drives.  It has a very nice temp sensor suite, too.  I’ve tried other utilities and they give about the same answer and similar results for HDD’s.  Pragmatically, I haven’t seen any real difference between SSD’s and HDD’s reliability beyond two decades or more of HDD use with 3-4 dead ones in that time.  Going on six? years of SSD’s with more devices than before (increased SSD availability and older kids who game) and am pretty happy with them.

Some Linux distros (ubuntu, maybe) have a drive stress/life left utility that gives similar info.  I’m looking for rough indicators more than exact values.

I’m definitely not a hardware pioneer; if  there had been some failures, probably back to hybrid HDD’s, almost as fast once they spin up.

Reply | Quote

Thanks. So 100% percent means “like brand new?” Or the opposite?

Ex-Windows user (Win. 98, XP, 7); since mid-2017 using also macOS. Presently on Monterey 12.15 & sometimes running also Linux (Mint).

MacBook Pro circa mid-2015, 15" display, with 16GB 1600 GHz DDR3 RAM, 1 TB SSD, a Haswell architecture Intel CPU with 4 Cores and 8 Threads model i7-4870HQ @ 2.50GHz.
Intel Iris Pro GPU with Built-in Bus, VRAM 1.5 GB, Display 2880 x 1800 Retina, 24-Bit color.
macOS Monterey; browsers: Waterfox "Current", Vivaldi and (now and then) Chrome; security apps. Intego AV

Reply | Quote

100% refers to the life remaining, so it’s like new.

Dell XPS 13/9310, i5-1135G7/16GB, KDE Neon 6.2
XPG Xenia 15, i7-9750H/32GB & GTX1660ti, Kubuntu 24.04
Acer Swift Go 14, i5-1335U/16GB, Kubuntu 24.04 (and Win 11)

1 user thanked author for this post.

OscarCP

Reply | Quote

OscarCP wrote:

All the information I have on how my Mac’s SSD is doing is from “About this Mac/System Report/Storage”

Have you tried the built-in First Aid?
6 Best Apps to Check Mac Hard Drive (or SSD) Health (anysoftwaretools.com)

cheers, Paul

Reply | Quote