SSD Write Endurance 25nm Vs 34nm

[Christopher]

He is an Ensign..... that is his job.

When this one eventually dies you should write to Samsung with the results of your little (maybe not so little) journey. Ask for a replacement because I promise you that a whole bunch of folks will be buying one of these drives as a result of your efforts. ...... They owe you that one replacement at least

will you repeat the test just to make sure it was not a lucky drive?

EDIT: Is there a sweepstake going on final figure?

Yes, the person who gets closest to the total bytes written receives a dead 830!

[Christopher]

Samsung 830 256GB Day 202

(GiB) 4,973,377
(TiB) 4,856
(PiB) 4.77


(Avg) 297.67 MB/s over the past 2000+ hours

(B1) Wear Leveling Count: 21,512

(B6) Erase Fail Count: 10
(B6) Runtime Bad Block Count: 11
(05) Reallocated Sectors: 45056

(POH) 4880


-------------------

Day 203 -- it's not dead yet.

[Dr. Vodka]

I have to say that I bought an 830 (my first SSD, and I'm blown away, to say the least) because of Christopher's testing. I know my little 64GB drive won't die because of some random reason, not anytime soon. It won't reach 4.5PB of writes, of course.. but it'll take a beating before saying goodbye. Great drives!


I hope the 840 series are as resilient as the 830s, if I decide to buy another SSD in a year or two... Thanks again guys for this epic thread and source of information. You all deserve a good round of beers

[Christopher]

I did test a 64GB 830 -- It went through a lot of PE cycles quickly. It's half as fast as the 256GB, but 1/4th the capacity. In this test it accumulated PE cycles very quickly.

--------------\
UPDATE
The 830 flagged another read failure, I fear we could be coming to the end of the 830's lifespan. The onset of read failures usually predict the end of testing -- the drive will start acting funky after more than a few.

[Anvil]

Status of the Intel 330 120GB

// nothing has changed, It's not running any tests. (still just idling)

831.75TB Host writes
2.81TiB Host reads
Reallocated sectors : 05 1184 +797 (387)
Available Reserved Space : E8 100 ""
MWI 1 (9)
[B5] Program fail count 563 (186)
[B6] Erase fail count 621 (201)
[F1] Total LBAs Written 27254740
[F2] Total LBAs Read 86201
[F9] Total NAND Writes 600556GB // ~586TiB
POH 2605 (1824)
MD5 --

---.--MiB/s on avg (hours)

--

Kingston SSDNow 40GB (X25-V)

1275.52TB Host writes 41796415*32)
Reallocated sectors : 05 82 +1
Available Reserved Space : E8 98
POH 11481
MD5 OK

33.14MiB/s on avg (~385 hours)

--

@christopher
The 830 is going to be a tough one to beat.

[Kain665]

Hmm... to Vertex 4 512GB or not to Vertex 4 512GB...

Just the sheer density of flash should see some good endurance, I'm not sure I want to pay that electricity bill though...

[Christopher]

Hmm... to Vertex 4 512GB or not to Vertex 4 512GB...

Just the sheer density of flash should see some good endurance, I'm not sure I want to pay that electricity bill though...

With a 512GB, it's more likely that one individual die will fail, rendering the drive a paperweight. The V4's controller supports redundancy but doesn't use it. You'd lose 64GB of capacity.

[Kain665]

Yeah I know, I'm just curious how long it will last. Decisions decisions...

[Christopher]

Yeah I know, I'm just curious how long it will last. Decisions decisions...

I have a 512GB 830.....

I bet I could get 10PB with it.

[Kain665]

I'm not taking any bets on how long the V4 may or may not last

[Anvil]

Status of the Intel 330 120GB

// Was restarted last night, been running for ~24 hours w/o issues.

842.10TB Host writes
2.84TiB Host reads
Reallocated sectors : 05 1219 +35 (387)
Available Reserved Space : E8 100 ""
MWI 1 (9)
[B5] Program fail count 580 (+17)
[B6] Erase fail count 639 (+18)
[F1] Total LBAs Written 27594058
[F2] Total LBAs Read 92971
[F9] Total NAND Writes 608074GB // ~594TiB
POH 3262
MD5 OK

124.86MiB/s on avg (~24 hours)

--

...meanwhile, the X25-V has been idling.

Kingston SSDNow 40GB (X25-V)

1276.43TB Host writes 41826023*32)
Reallocated sectors : 05 82
Available Reserved Space : E8 98
POH 11515
MD5 --

--.--MiB/s on avg (~-- hours)

It's been running for a few minutes, will check later tonight.

--

Working on updating the charts...

[Anvil]

Updated charts

Endurance_cr_20121010.png

Endurance_PEsummary_20121010.png

I've included the SanDisk Extreme and the Intel X25-M G1 this time. (next update they are gone)

Endurance_failed_drives.png

There are a few drives that needs to be sorted out, the Vertex 4 128GB and the Samsung 830 64GB.
(just point me to the latest post or a short summary for the status in post 1)

Let me know if there's anything I've forgotten!

-

[B.A.T]

Here are a long due update :

m4
1746.0205 TiB
7182 hours
Avg speed 74.44 MiB/s.
AD 147 to 123
P/E 30056
C3 1150608 to 3511415
01 383 to 1044 (61 to 1)
CE 90
MD5 OK.
Reallocated sectors : 00


Intel X25-E 64GB
1.10 PiB
3647-30=3617 hours
Reallocated sectors : 6 to 115
Available Reserved space: 84 to 81
MWI= 92
MD5 =OK
81.37 MiB/s on avg


Mtron Pro 7025 32GB
Paused until new firmware is updated.

OCZ Vertex 1 120GB
Will start after an update to the latest firmware.

[Christopher]

Samsung 830 256GB Day 207

(GiB) 5,079,604
(TiB) 4,960
(PiB) 4.87


(Avg) 297.87 MB/s over the past 2000+ hours

(B1) Wear Leveling Count: 21,969

(B6) Erase Fail Count: 11
(B6) Runtime Bad Block Count: 13
(05) Reallocated Sectors: 53248

(POH) 4980


-------------------

We're close -- very close -- to 5PiB. No other changes, though.

[B.A.T]

Todays update :

m4
1762.1309 TiB
7246 hours
Avg speed 74.42 MiB/s.
AD 123 to 113
P/E 30331
C3 3511415 to 5566191
01 1044 to 1543 (1)
CE 90
MD5 OK.
Reallocated sectors : 00


Intel X25-E 64GB
1.12 PiB
3710-30=3680 hours
Reallocated sectors : 115 to 156
Available Reserved space: 81 to 79
MWI= 92
MD5 =OK
81.07 MiB/s on avg


Mtron Pro 7025 32GB
Paused until new firmware is updated.

OCZ Vertex 1 120GB
Will start after an update to the latest firmware.

[Anvil]

Status of the Intel 330 120GB

// Is no longer detected,
(I've tried different pc's as well as leaving it disconnected overnight)

It looks to have disappeared sometime the 12th, I'll find out in the log-file.
It took down all drives so nothing was written for a couple of days.

--

Kingston SSDNow 40GB (X25-V)

1292.80TB Host writes 42362583*32)
Reallocated sectors : 05 768 (82)
Available Reserved Space : E8 86 (98)
POH 11656
MD5 OK

32.36MiB/s on avg (~19 hours)

It's starting to show some wear, it's still got a bit of reserved space left...

[Anvil]

Final stats for the Intel 330 120GB

863.33TB Host writes
Reallocated sectors : 05 1464
Available Reserved Space : E8 100
MWI 1
[B5] Program fail count 694
[B6] Erase fail count 770
[F9] Total NAND Writes ~609.3TiB
POH n // I'll find out

Intel_330_final_stats.png

[B.A.T]

R.I.P Intel 330.
600 TiB of nand write isn't a lot for a 120GB drive.

Here are todays update:
m4
1777.9380 TiB
7307 hours
Avg speed 74.45 MiB/s.
AD 113 to 104
P/E 30600
C3 5566191 to 8457546
01 1543 to 2285 (1)
CE 90
MD5 OK.
Reallocated sectors : 00


Intel X25-E 64GB
1.14 PiB
3772-30=3742 hours
Reallocated sectors : 156 to 207
Available Reserved space: 79 to 78
MWI= 92
MD5 =OK
81.07 MiB/s on avg


Mtron Pro 7025 32GB
Paused until new firmware is updated.

OCZ Vertex 1 120GB
Will start after an update to the latest firmware.

[Christopher]

Samsung 830 256GB Day 211

(GiB) 5,204,510
(TiB) 5,082
(PiB) 4.99


(Avg) 298.17 MB/s over the past 2000+ hours

(B1) Wear Leveling Count: 22,507

(B6) Erase Fail Count: 12
(B6) Runtime Bad Block Count: 15
(05) Reallocated Sectors: 61440

(POH) 5098


-------------------

A few more erase fails, but 5PiB isn't that far away.

[Anvil]

Kingston SSDNow 40GB (X25-V)

1295.32TB Host writes 42444949*32)
Reallocated sectors : 05 768 -
Available Reserved Space : E8 86 (86)
POH 11679
MD5 OK

31.95MiB/s on avg (~42 hours)

--

It's time to start looking for a new drive to test.

[Christopher]

Samsung 830 256GB Day 211

(GiB) 5,221,720
(TiB) 5,099
(PiB) 5.01


(Avg) 298.17 MB/s over the past 2000+ hours

(B1) Wear Leveling Count: 22,581

(B6) Erase Fail Count: 12
(B6) Runtime Bad Block Count: 15
(05) Reallocated Sectors: 61440

(POH) 5115


-------------------

That's 5PiB!

[Kain665]

Double post.

[Kain665]

Hey guys, quick question. What do you think of the health of this drive? SF2282 with Intel Async 25nm NAND. Raw Read Error rate fluctuates from time to time, I've seen it as low as 103. Only 3 retired blocks, and no other obvious issues.

D2C480A1.png

[canthearu]

Drive looks fine to me. Only 2 reallocated blocks, which is nothing for the SSD.

As long as the normalised value for raw error rate stays above 50, the drive isn't failing in that test either.

[Christopher]

I haven't stopped the 830 in nearly five months, but I'm thinking about pulling it down and running some tests on it before resuming testing.

There are lots of questions I have about its performance after 5PiB. It's already got 5.3million GiB on it.