The SSD Endurance Experiment: Finally, They Are All Dead

[BiG StroOnZ]

More than 18 months ago, I vowed to push all six drives to their bitter ends. I didn't do so in the name of god or country or even self-defense, either. I did it just to watch them die.

 

I slowly squeezed out every last drop of life with a relentless stream of writes far more demanding than anything the SSDs would face in a typical PC. To make matters worse, I exploited their suffering by chronicling the entire process online.

 

Today, that story draws to a close with the final chapter in the SSD Endurance Experiment. The last two survivors met their doom on the road to 2.5PB, joining four fallen comrades who expired earlier. It's time to honor the dead and reflect on what we've learned from all the carnage.

 

This experiment sought to find out by writing a near-constant stream of data to Corsair's Neutron GTX 240GB, Intel's 335 Series 240GB, Kingston's HyperX 3K 240GB, Samsung's 840 Series 250GB, and Samsung's 840 Pro 256GB.

 

All of the drives surpassed their official endurance specifications by writing hundreds of terabytes without issue. Delivering on the manufacturer-guaranteed write tolerance wouldn't normally be cause for celebration, but the scale makes this achievement important. Most PC users, myself included, write no more than a few terabytes per year. Even 100TB is far more endurance than the typical consumer needs.

 

After receiving a black mark on its permanent record, the 840 Series sailed smoothly up to 800TB. But it suffered another spate of uncorrectable errors on the way to 900TB, and it died without warning before reaching a petabyte. 

 

Intel's 335 Series failed much earlier, though to be fair, it pulled the trigger itself. The drive's media wear indicator ran out shortly after 700TB, signaling that the NAND's write tolerance had been exceeded. Intel doesn't have confidence in the drive at that point, so the 335 Series is designed to shift into read-only mode and then to brick itself when the power is cycled.

 

The reaper came for the Kingston HyperX 3K next. As with the 335 Series, the SMART data's declining life indicator foretold the drive's death and triggered messages warning that the end was nigh. The flash held up nicely through 600TB, but it suffered a boatload of failures and reallocated sectors leading up to 728TB, after which it refused to write.

 

The next failure occurred after the 840 Series bit the dust. Corsair's Neutron GTX was practically flawless through 1.1PB—that's petabytes—but it posted thousands of reallocated sectors and produced numerous warning messages over the following 100TB.  

 

The SSD Endurance Experiment represents the longest test TR has ever conducted. It's been a lot of work, but the results have also been gratifying. Over the past 18 months, we've watched modern SSDs easily write far more data than most consumers will ever need. Errors didn't strike the Samsung 840 Series until after 300TB of writes, and it took over 700TB to induce the first failures. The fact that the 840 Pro exceeded 2.4PB is nothing short of amazing, even if that achievement is also kind of academic.

 

The important takeaway is that all of the drives wrote hundreds of terabytes without any problems. Their collective endurance is a meaningful result.

 

 

 

 

Source: http://techreport.com/review/27909/the-ssd-endurance-experiment-theyre-all-dead

 

This is a pretty exceptional test, seeing that most of these drives wrote hundreds of terabytes before failing, some even reaching a petabyte. Really shows how reliable SSD's actually are in today's age, as long as you go with a decent company. Only gripe is I wish they had a Crucial drive in the test, as I believe they are pretty big in the SSD game. 

[H4X3R]

I agree. A crucial ssd would be a nice addition to the comparison.

[Enderman]

YES FINALLY i wanted to see how long they lasted...

 

now they need to start doing 3d nand drives like the 850 and the upcoming intel SSDs

[MLGNOSCOPEHEADSHOTGETREKT]

And everyone says newtron ssd's are bad

#CORSAIR

[Eroda]

so intel basically set their limits for 100% safe data then make sure you cannot surpass that.. if thats the case cant really see an issue with that

[Trik'Stari]

WOOO HOOO! 840 pro got up to 2.4Pb!

[Tim Drake]

I knew I should of just saved up for an 840 EVO :'(

[Sakkura]

Should put the nail in the coffin on the SSD endurance fears you still hear about here and there.

 

But it would be nice if SSD makers could get their drives to wear out a little more elegantly. Intel's solution was as expected, and honestly I would rather get a consistent, controlled shutdown at 700TB than a sudden, total failure at 2PB+. Not that it really matters to most of us anyway.

[Thony]

Nice test, but how much do I use per day is my question

Does anybody know about an application that could monitor the data written to a specific dry from when I push the button on my pc to when I'm about to shut it down for the night ?

[SurvivorNVL]

I wonder how the endurance of NVME / NVMEHCI drives will be once they're out for consumers.  Maybe we'll hit a few exabytes

[connorpiper]

It would be nice to have this done with Samsungs 3D NAND, and ADATA.

[bcredeur97]

I knew I should of just saved up for an 840 EVO :'(

the evo wasnt tested. even if it was i dont think it would of got past 1.5PB

 

Reallistically your MLC nand MX100 should be fine for a Petabyte.

[Tim Drake]

the evo wasnt tested. even if it was i dont think it would of got past 1.5PB

 

Reallistically your MLC nand MX100 should be fine for a Petabyte.

Whoops

[PopReference]

After 2+ years my Kingston 3k only has 5.4TB in 14161 as an OS, that has taken some abuse. Think I'm less worried now.

[Octagoncow]

I doubt they're going to do it; but, it would be cool to redo this with the current generation of those SSDs now (ex Samsung 850's vs 840's). It would be interesting to see if the endurance changes.

[Ertman]

wow corsair.

[werto165]

After 2+ years my Kingston 3k only has 5.4TB in 14161 as an OS, that has taken some abuse. Think I'm less worried now.

I've written 14.5TB in 1.5 years, but from the looks of these it's fine

 

 

 1: (SSD Raw Read Error Rate)               Normalized Rate:  95  Sectors Read: 6259485 Read Errors: 0

 5: (SSD Retired Block Count)               Spare blocks remaining 100%  Retired Block 0

 9: (SSD Power-On Hours)                    Value 95  Total 5222 hrs 20 mins

12: (SSD Power Cycle Count)                 Power Cycle Life Remaining  99%  Number of power cycles 1038

171: (SSD Program Fail Count)                Program Error Count 0

172: (SSD Erase Fail Count)                  Erase Error Count 0

174: (SSD Unexpected power loss count)       Unexpected power loss Count 113

177: (Wear Range Delta)                      Wear Range Delta 3%

181: (Program Fail Count)                    Program Error Count 0

182: (Erase Fail Count)                      Erase Error Count 0

187: (SSD Reported Uncorrectable Errors)     Normalized Value 100  lifetime URAISE Errors 0

189: (Unrecognized Attribute)                Value:  29 Raw Data: 1d 00 4b 00 0c 00 00

194: (SSD Temperature Monitoring)            Normalized temp 29   Current 29  High 75 Low 12

195: (SSD ECC On-the-fly Count)              Normalized Value 120   Sectors Read 6259485  UECC Count 0

196: (SSD Reallocation Event Count)          Normalized Value 100   Reallocation Event Count 0

201: (SSD Uncorrectable Soft Read Error Rate)Normalized Value 120  Sectors Read 6259485  Uncorrectable Soft Error Count 0

204: (SSD Soft ECC Correction Rate (RAISE)   Normalized Value 120  Sectors Read 6259485  Soft ECC Correction Count 0

230: (SSD Life Curve Status)                 Normalized Value 100

231: (SSD Life Left)                         Life Remaining 100%

233: (SSD Internal Reserved)                 17000

234: (SSD Internal Reserved)                 14355

241: (SSD Lifetime writes from host)         lifetime writes   14355

242: (SSD Lifetime reads from host)          lifetime reads   13116

 

No signs of premature failure. 

[Biggerisbetter]

Absolutely brilliant, thanks that post OP. I feel that samsung pro SSD's are the best around. They do cost more upfront, but I believe they can be passed down from build to build. Although this was not a very scientific experiment. It was very informative. I wish more channels would run experiments like this. Thx again OP

[Sharkyx1]

I knew I should of just saved up for an 840 EVO :'(

840 EVOs are nothing but trouble, rave reviews out the gate, but a year down the road they've slowed down to less than mechanical speeds

[Sharkyx1]

 

I've written 14.5TB in 1.5 years, but from the looks of these it's fine

 

My 840 Evo has 38.16 TB of writes according to samsung magician and i have no idea why

[werto165]

My 840 Evo has 38.16 TB of writes according to samsung magician and i have no idea why

I guess it depends on how full you keep your drive, as far as I know SSD's try to "split" the load over multiple chips rather than just occupying one section of NAND. 

[Sharkyx1]

I guess it depends on how full you keep your drive, as far as I know SSD's try to "split" the load over multiple chips rather than just occupying one section of NAND. 

90 Gigs free now, but its been more full before, i cleaned up quite a bit and moved downloads and images to my HDD

[killertek78]

My 840 Evo has 38.16 TB of writes according to samsung magician and i have no idea why

Swap/hibernate files?

[spartaman64]

I knew I should of just saved up for an 840 EVO :'(

they tested the 840 pro the 840 evo is using TLC nand so it should fail fairly early your crucial ssd with its MLC nand should last longer

[Sakkura]

I doubt they're going to do it; but, it would be cool to redo this with the current generation of those SSDs now (ex Samsung 850's vs 840's). It would be interesting to see if the endurance changes.

 

3D NAND would make it take ages to complete. I mean this experiment already took over a year, with an 850 Pro it could take the rest of the decade.