SSD Write Endurance 25nm Vs 34nm

[Christopher]

I'm not certain what is going on with the M3P, but I'm not physically there at the moment to go over it more thoroughly. However, I was just taking some screenshots of ASU and CDM, and the M3P's ASU instance just "unfroze"

IT LIVES! Sort of, it appears ASU just locks up periodically, and for hours, then it continues for a short time.

UPDATE,
I force-quite ASU, then tried reading and writing some files to the M3P, and all was well. I started ASU back up, and now it won';t make it past the spash, and now Windows cant get at the drive either it seems.

[canthearu]

Sandisk Extreme G25 - 120gb - Day 31

Drive hours: 715
ASU GiB written: 317,241.37 GiB (309.81 TiB)
Avg MB/s: 125.28 MB/s (176.77 hours)
MD5: OK (verified off disk now)

Host GB written (F1): 319,096 GiB (311.58 TiB)
NAND writes (E9): 238,043 GiB (232.46 TiB)

Retired Block Count (05): 4 raw
Failure count (AB, AC): 1 program, 1 erase
Raw Error Rate (01): 117 normalized
Media Wearout Indicator (E7): 37 normalized
Wear Range Delta (B1): 4 raw
Reported Uncorrectable Errors (BB): 0 raw

[B.A.T]

Todays update:

m4
1160.2395 TiB
4853 hours
Avg speed 73.71MiB/s.
AD 208 to 202
P/E 19988.
C3 4312
CE 58
MD5 OK.
Reallocated sectors : 00


Intel X25-M G1 80GB
766.49 TiB
24141 hours
Reallocated sectors : 151 to 161
Available Reserved space: 60 to 59
MWI= 1 to 255 to 252
MD5 =OK
34.06 MiB/s on avg


Intel X25-E 64GB
425.94 TiB
1322-30=1292 hours
Reallocated sectors : 0
Available Reserved space: 100
MWI= 97
MD5 =OK
91.78 MiB/s on avg


Mtron Pro 7025 32GB
ASU write 88.0086 TiB
Host write 165,9422 TiB
Total host write 213,4778 TiB
Day 33
MWI= 94 to 93
MD5 =OK
47.95 MiB/s on avg


No big changes since last update. The Mtron dropped out 3 times in this period but runs normal now.

[Christopher]

I'm traveling today, but I'm unable to reach the endurance rig. I think something terrible has happened, and I'm not there to see about it.

I feel like a negligent parent. I suspect the M3P locked the rig up.

[Kain665]

I'm calling child support!

[Anvil]

Intel 330 120GB

428.57TB Host writes
1.60TiB Host reads
Reallocated sectors : 05 0
Available Reserved Space : E8 100
MWI 51
[F1] Total LBAs Written 14043401
[F2] Total LBAs Read 52304
[F9] Total NAND Writes 309363GB
POH 1055
MD5 OK

124.41MiB/s on avg (~68 hours)

--

Kingston SSDNow 40GB (X25-V)

1095.37TB Host writes (35893193*32)
Reallocated sectors : 05 52
Available Reserved Space : E8 99
POH 9788
MD5 OK

33.84MiB/s on avg (~68 hours)

--

Endurance_cr_20120707.png

The 330 is ahead of the X25-E on Host writes , it won't last though.

The Plextor stays on the chart until we get a new/final notice about it's health.

[B.A.T]

Todays update:

m4
1169.2421 TiB
4889 hours
Avg speed 73.63 MiB/s.
AD 202 to 197
P/E 20142.
C3 4312
CE 58
MD5 OK.
Reallocated sectors : 00


Intel X25-M G1 80GB
770.51 TiB
24176 hours
Reallocated sectors : 161 to 170
Available Reserved space: 59 to 58
MWI= 252 to 242
MD5 =OK
33.59 MiB/s on avg


Intel X25-E 64GB
437.27 TiB
1357-30=1327 hours
Reallocated sectors : 0
Available Reserved space: 100
MWI= 97
MD5 =OK
91.87 MiB/s on avg


Mtron Pro 7025 32GB
ASU write 90.8840 TiB
Host write 171,4600 TiB
Total host write 218,9956 TiB
Day 34
MWI= 93
MD5 =OK
47.87 MiB/s on avg


Only changes is that the Mtron dropped out 1 time since last update.

[Christopher]

I just got home, looking into the Plextor now. The system is indeed on but locked up, so I suspect the M3P is truly dead.

I removed the M3P, and the rig unfroze. The Samsung is still chugging along with 2600.36TiB.

I'll run some tests on the M3P now

-------------------------------------
The M3P was put into another system. . Here is some SMART info, reserved space is at 0. I have ASU running at the moment, but for how long, I'm not certain.

The M3P was uninitialized when I put it in the other system to check it out, so all ASU data is gone.




I believe MWI is at -20 as well.



Plextool issued a warning as well when I ran it on another system.

Interestingly, when I can get benchmarks to run (though not complete) seq write speed is up 30MB/s over new! That's kinda crazy, but seq. read is diminished by an equal amount.

[Christopher]

Samsung 830 256GB Day 110

(GiB) 2,668,510.29
(TiB) 2,605.96

(Avg) 281.96MB/s

(B1) Wear Leveling Count: 11,569

(B6) Erase Fail Count: 9
(05) Reallocated Sectors: 36864
(POH) 2659

-------------------
The reduction in Avg speed is due to the couple of days the system spent locked up I think, not real sure.

Incidentally, the PlexTool warning is priceless (so is the dialog footer "Plextool gives you a information" -- love it)

[canthearu]

Sandisk Extreme G25 - 120gb - Day 33

Drive hours: 795
ASU GiB written: 341,756.81 GiB (333.75 TiB)
Avg MB/s: 125.20 MB/s (232.59 hours)
MD5: OK (verified off disk now)

Host GB written (F1): 343,703 GiB (335.65 TiB)
NAND writes (E9): 256,441 GiB (250.43 TiB)

Retired Block Count (05): 4 raw
Failure count (AB, AC): 1 program, 1 erase
Raw Error Rate (01): 118 normalized
Media Wearout Indicator (E7): 32 normalized
Wear Range Delta (B1): 3 raw
Reported Uncorrectable Errors (BB): 0 raw

[Anvil]

Intel 330 120GB

439.80TB Host writes
1.62TiB Host reads
Reallocated sectors : 05 0
Available Reserved Space : E8 100
MWI 49
[F1] Total LBAs Written 14411251
[F2] Total LBAs Read 53212
[F9] Total NAND Writes 317466GB
POH 1081
MD5 OK

124.51MiB/s on avg (~94 hours)

--

Kingston SSDNow 40GB (X25-V)

1098.36TB Host writes (35991195*32)
Reallocated sectors : 05 54
Available Reserved Space : E8 99
POH 9814
MD5 OK

33.66MiB/s on avg (~94 hours)

--

[B.A.T]

Todays update:

m4
1175.1555 TiB
4912 hours
Avg speed 73.61 MiB/s.
AD 197 to 194
P/E 20244.
C3 4312
CE 58
MD5 OK.
Reallocated sectors : 00


Intel X25-M G1 80GB
773.27 TiB
24200 hours
Reallocated sectors : 170 to 176
Available Reserved space: 58
MWI= 242 to 235
MD5 =OK
33.71 MiB/s on avg


Intel X25-E 64GB
444.70 TiB
1381-30=1351 hours
Reallocated sectors : 0
Available Reserved space: 100
MWI= 97
MD5 =OK
91.88 MiB/s on avg


Mtron Pro 7025 32GB
ASU write 90.8840 TiB
Host write 176,9984 TiB
Total host write 224,5339 TiB
Day 35
MWI= 93
MD5 =OK
47.83 MiB/s on avg


And....Surprise! The Mtron dropped out again since last update.

[Christopher]

Samsung 830 256GB Day 111

(GiB) 2,693,661.29
(TiB) 2,630.53

(Avg) 282.46MB/s

(B1) Wear Leveling Count: 11,677

(B6) Erase Fail Count: 9
(05) Reallocated Sectors: 36864
(POH) 2683

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

[Anvil]

Intel 330 120GB

450.25TB Host writes
1.74TiB Host reads
Reallocated sectors : 05 0
Available Reserved Space : E8 100
MWI 48
[F1] Total LBAs Written 14753848
[F2] Total LBAs Read 57108
[F9] Total NAND Writes 325019GB
POH 1106
MD5 OK

124.68MiB/s on avg (~24 hours)

--

Kingston SSDNow 40GB (X25-V)

1101.35TB Host writes (36088961*32)
Reallocated sectors : 05 54
Available Reserved Space : E8 99
POH 9839
MD5 OK

35.45MiB/s on avg (~24 hours)

--

Both drives were lightly cleaned last night, by using ASU's built in TRIM trigger.

[B.A.T]

Todays update:

m4
1181.2778 TiB
4936 hours
Avg speed 73.59 MiB/s.
AD 194 to 190
P/E 20348.
C3 4312
CE 58
MD5 OK.
Reallocated sectors : 00


Intel X25-M G1 80GB
775.96 TiB
24224 hours
Reallocated sectors : 176 to 188
Available Reserved space: 58 to 57
MWI= 235 to 226
MD5 =OK
33.45 Mi/s on avg


Intel X25-E 64GB
452.34 TiB
1405-30=1375 hours
Reallocated sectors : 0
Available Reserved space: 100
MWI= 97
MD5 =OK
91.79 MiB/s on avg


Mtron Pro 7025 32GB
ASU write 96.0459 TiB
Host write 181,3895 TiB
Total host write 228,9250 TiB
Day 36
MWI= 93
MD5 =OK
47.89 MiB/s on avg


And....again the Mtron dropped out. Runs fine now.

[canthearu]

Sandisk Extreme G25 - 120gb - Day 35

Drive hours: 836
ASU GiB written: 359,546.25 GiB (351.12 TiB)
Avg MB/s: 125.26 MB/s (272.85 hours)
MD5: OK (verified off disk now)

Host GB written (F1): 361,577 GiB (353.10 TiB)
NAND writes (E9): 269,789 GiB (263.47 TiB)

Retired Block Count (05): 4 raw
Failure count (AB, AC): 1 program, 1 erase
Raw Error Rate (01): 118 normalized
Media Wearout Indicator (E7): 28 normalized
Wear Range Delta (B1): 3 raw
Reported Uncorrectable Errors (BB): 0 raw

[Christopher]

Samsung 830 256GB Day 112

(GiB) 2,727,423.40
(TiB) 2,663.50

(Avg) 301.37MB/s

(B1) Wear Leveling Count: 11,823

(B6) Erase Fail Count: 9
(05) Reallocated Sectors: 36864
(POH) 2715

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

[Anvil]

About 5 hours ago the 330 reported 2 reallocated sectors

B5 and B6 both changed from 0 to 1. (not sure what they are, will check up, both are listed as unknown)

--

Intel 330 120GB

460.54TB Host writes
1.77TiB Host reads
Reallocated sectors : 05 2
Available Reserved Space : E8 100
MWI 47
[B5] 1
[B6] 1
[F1] Total LBAs Written 15090825
[F2] Total LBAs Read 57959
[F9] Total NAND Writes 332438GB
POH 1130
MD5 OK

124.74MiB/s on avg (~48 hours)

--

Kingston SSDNow 40GB (X25-V)

1104.09TB Host writes (36178803*32)
Reallocated sectors : 05 54
Available Reserved Space : E8 99
POH 9863
MD5 OK

34.34MiB/s on avg (~48 hours)

[B.A.T]

Todays update:

m4
1186.9700 TiB
4959 hours
Avg speed 73.56 MiB/s.
AD 190 to 187
P/E 20446.
C3 4312 to 4457 (after firmware update)
CE 58
MD5 OK.
Reallocated sectors : 00


Intel X25-M G1 80GB
778.54 TiB
24247 hours
Reallocated sectors : 188 to 199
Available Reserved space: 57
MWI= 226 to 219
MD5 =OK
33.41 Mi/s on avg


Intel X25-E 64GB
459.46 TiB
1428-30=1375 hours
Reallocated sectors : 0
Available Reserved space: 100
MWI= 97 to 96
MD5 =OK
91.73 MiB/s on avg


Mtron Pro 7025 32GB
ASU write 99.3457 TiB
Host write 187,7499 TiB
Total host write 235,2854 TiB
Day 37
MWI= 93
MD5 =OK
47.83 MiB/s on avg


Changes since last update:
-The Mtron dropped out 1 time. I'll be pausing the endrurance test on the Mtron until Iæll figured out how to update firmware on it. I need to find some documentation first.
- I'll updatet to the lastest firmware on the m4 to avoid the 5K bug. The update went well and the drive runs normal again. Only change was a smal increase in Cumulative Corrected ecc because it was without power for aprox 1 hour.


@Anvil
Can you update the first post with the new firmware on the m4?

[Anvil]

@B.A.T

Will do

@christopher

What's the final verdict on the Plextor?

--

Endurance_cr_20120710.png

[Christopher]

The M3P is dead. I couldn't get it running for more than a few moments; It's quite curious actually.

However, today I tried the drive out, and all the data was there, I grabbed some more smart data (a lot more reallocations were seen vs. the last data I posted). Now the drive can be detected by the UEFI, but no dice. It will hang the system now, so looks like it's all gone.

Here is the data I grabbed before death:




896.699 TiB in host writes. I did manage to figure out what all of the smart attributes are though, so that's a plus.

I really liked this drive, it's sad to see it go. It probably used over 7800 PE cycles, and while I was hoping for more, the drive is just way to damn fast.

[AMD_Freak]

sent my contact at Plextor the above info and 3 days later I get this reply

Hi *****,
Plextor SSDs have been through testings by our RDs. For heavy users, the maximum could reach to 20G/ day. It can write about 36.5 TB/ 5 years. TiB written 856 could be used for 115 years.
- Plextor

I also linked then the thread, I'm quite disappointed in the short time it took to get the drive to fail it did go fast while it was kicking.

[canthearu]

I really liked this drive, it's sad to see it go. It probably used over 7800 PE cycles, and while I was hoping for more, the drive is just way to damn fast.

I know it is disappointing to see a drive die early ... but it did do twice the specified P/E cycles of the NAND.

So a fairly average result ... but not bad at all in real world terms

[Anvil]

I wouldn't call it disappointing, shrinking leads to fewer P/E cycles, binning/QC can make a difference.

MWI behaves differently from other drives though, could be that this is just as good as being based on "P/E count".

I'd say it did well!

--

The 330 lost 2 more sectors, current count is up from 2 to 4. (happened during the night)
B5 and B6 looks to go hand in hand with 05 and are both up to 2.

[A||uSiOn]

I actually was speaking to a intel rep and showed them this thread and they said they would closely follow this thread and investigate the quick failure rate of the 520 series drive.