SSD Write Endurance 25nm Vs 34nm

[canthearu]

Interesting, the drive locked up again, so I secure erased it this time. Now smart reports:



Never seen a sandforce drive resist outright panic locking, and never seen 0% from a sandforce drive in SMART before.

And no, the drive is not readonly, but so far, no data loss either.

[Christopher]

^^ 750GB total nand writes, Plextor did get premium nands from Toshiba then for extra ~100GB writes then lol..

Yeah, but the Plexty wrote 20,000GB a day... That's nearly 190PE cycles a day. If it was doing 70MB/s instead of 270MB/s, it would have gone a lot further.

[B.A.T]

Here are todays update:

m4
1587.1365 TiB
6554 hours
Avg speed 75.06 MiB/s.
AD 219 to 214
P/E 27321
C3 28392 to 38972
01 11 to 15 (99)
CE 90
MD5 OK.
Reallocated sectors : 00


Intel X25-E 64GB
951.26 TiB
3020-30=2990 hours
Reallocated sectors : 44 to 58
Available Reserved space: 90
MWI= 93
MD5 =OK
86.20 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.

[B.A.T]

Here are todays update:

m4
1596.8173 TiB
6591 hours
Avg speed 75.04 MiB/s.
AD 214 to 208
P/E 27486
C3 38972 to 55988
01 15 to 23 (99 to 98)
CE 90
MD5 OK.
Reallocated sectors : 00


Intel X25-E 64GB
962.26 TiB
3058-30=3028 hours
Reallocated sectors : 58 to 72
Available Reserved space: 90
MWI= 93
MD5 =OK
86.01 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 181

(GiB) 4,445,851
(TiB) 4,341
(PiB) 4.26


(Avg) 297.60 MB/s over the past 1600+ hours

(B1) Wear Leveling Count: 19,233

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

(POH) 4372


-------------------
We're closing in on 20K PE cycles, it should happen around 4,620,000 GiB.

[canthearu]

Ok, I've been a bit caught up building my hackintosh over the last couple of days, so I need to report the death of the old sandisk 120gig

It locked up again, but never recovered:

Final Stats:

Hours: 2332 (97 days)

Host GB written (F1): 1,006,659 GiB (983.07 TiB)
NAND writes (E9): 751,782 GiB (734.16 TiB)

Retired Block Count (05): 4 raw
Failure count (AB, AC): 1 program, 1 erase
Raw Error Rate (01): 120 normalized
Media Wearout Indicator (E7): 0 normalized (failed)
Wear Range Delta (B1): 4 raw
Reported Uncorrectable Errors (BB): 0 raw

So that is it ... the sandisk lived and died in much to same way as the corsiar force 3 did ... so 24nm Samsung NAND seems about as good as 25nm Micron NAND.

Bit of a fail that it didn't reach 1PiB written, oh well.

For now, I won't be endurance testing any drives ... maybe when 19nm NAND is more common.

[kensiko]

RIP sandisk.

At least it was over the specified PE/count.

By the way, how can it be possible for the Chronos Deluxe to stay at 350$ for the 480GB?
Isn't it supposed to contains high priced 32nm Toggle-mode flash?

[Zaxx]

The new Chronos Deluxe with a 'DX' in the model number doesn't use 32nm toggle. They are sorta pulling an Ocz...keeping the same model name from a previously fast drive for a newer drive with newer hardware and slower speeds. You can probably expect the faster 'old' chronos deluxe to start disappearing before long or IF they sell enuff of them, they'll keep 'em around like Ocz did with the V3 MaxIOPs line but stay with the premium price ofc.

[Christopher]

There are a couple excellent drives coming out soon that I'm looking forward to, so maybe one of those will become fodder for testing.

The 512/480 Mushkin is using IMFT, which is probably a good thing. The 2281s with 32nm Toggle are nothing to write home about (at 480GB). With the 2282, that changes I think.

[B.A.T]

R.I.P Sandisk. This thread has claimed another victim.

Here are todays update:

m4
1607.6699 TiB
6634 hours
Avg speed 75.00 MiB/s.
AD 208 to 202
P/E 27672
C3 55988 to 70750
01 23 to 28 (98 to 97)
CE 90
MD5 OK.
Reallocated sectors : 00


Intel X25-E 64GB
975.44 TiB
3103-30=3073 hours
Reallocated sectors : 72 to 88
Available Reserved space: 90 to 89
MWI= 93
MD5 =OK
85.83 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.

[zads]

Interesting, the drive locked up again, so I secure erased it this time. Now smart reports:



Never seen a sandforce drive resist outright panic locking, and never seen 0% from a sandforce drive in SMART before.

And no, the drive is not readonly, but so far, no data loss either.

You ran out of spare blocks.

[zads]

No, because secure erase would only erase the NAND that was currently used on the drive. If none of the NAND is allocated, then secure erase wouldn't do much.

Depending on the controller, Secure erase doesn't erase the nand either.

[B.A.T]

Here are todays update:

m4
1620.4985 TiB
6684 hours
Avg speed 74.92 MiB/s.
AD 202 to 195
P/E 27890
C3 70750 to 99650
01 28 to 41 (97 to 96)
CE 90
MD5 OK.
Reallocated sectors : 00


Intel X25-E 64GB
989.15 TiB
3150-30=3120 hours
Reallocated sectors : 88 to 109
Available Reserved space: 89 to 88
MWI= 93
MD5 =OK
85.59 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.

[B.A.T]

Quiet days in here now....

Here are todays update:

m4
1627.4046 TiB
6711 hours
Avg speed 74.91 MiB/s.
AD 195 to 191
P/E 28007
C3 99650 to 128090
01 41 to 51 (96 to 95)
CE 90
MD5 OK.
Reallocated sectors : 00


Intel X25-E 64GB
996.90 TiB
3177-30=3147 hours
Reallocated sectors : 109 to 123
Available Reserved space: 88
MWI= 93
MD5 =OK
85.45 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.

[3rase]

Anyone thinking of testing the Corsair Neutron?

[Christopher]

Samsung 830 256GB Day 187

(GiB) 4,590,570
(TiB) 4,482
(PiB) 4.40


(Avg) 297.59 MB/s over the past 1600+ hours

(B1) Wear Leveling Count: 19,859

(B6) Erase Fail Count: 10
(05) Reallocated Sectors: 40960

(POH) 4511


-------------------
FINALLY!!!

A new erase failure!!

It's been over 2PiB since the last. It might have been more like 3PB since the last failure. I hope they start occurring with more frequency; the 830 is really boring me to death. It's like it doesn't even care that it's writing most of the time.

[Computurd]

Depending on the controller, Secure erase doesn't erase the nand either.

can you expand on that a bit? i know that SE involves a different voltage being applied to the nand to clear it, but i assumed that would erase it as well. I have heard elsewhere that isnt always the case, but not an explanation of how it works.

[B.A.T]

Here are todays update:

m4
1637.4888 TiB
6750 hours
Avg speed 74.90 MiB/s.
AD 191 to 185
P/E 28178
C3 128090 to 171093
01 51 to 67 (95 to 93)
CE 90
MD5 OK.
Reallocated sectors : 00


Intel X25-E 64GB
1008.140 TiB
3216-30=3186 hours
Reallocated sectors : 123 to 147
Available Reserved space: 88 to 87
MWI= 93
MD5 =OK
85.45 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 188

(GiB) 4,590,570
(TiB) 4,482
(PiB) 4.40


(Avg) 297.59 MB/s over the past 1600+ hours

(B1) Wear Leveling Count: 20,017

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

(POH) 4547


-------------------
There was another block failure, except this time I think it's a read failure. There are smart attributes for erase and program failures, but not read failures. 11 total reallocation events were comprised of 10 erase failures, 0 program, leaving the remaining one which occured today a read failure based on process of elimination.

[kensiko]

Next test should be done on the Samsung 840 TLC!

http://www.anandtech.com/show/6329/s...-based-840-ssd

The life span is supposed to be 10 times less.

[Christopher]

I'll admit, I'm interested in checking it out. I almost can't believe they're using TLC, but what the hell? I'm game. I thought my 830 might have been spiraling downward, but it looks like it won't be dead any time soon.

[B.A.T]

Here are todays update:

m4
1646.0629 TiB
6783 hours
Avg speed 74.91 MiB/s.
AD 185 to 180
P/E 28324
C3 171093 to 227775
01 67 to 83 (93 to 92)
CE 90
MD5 OK.
Reallocated sectors : 00


Intel X25-E 64GB
1017.66 TiB
3250-30=3220 hours
Reallocated sectors : 147 to 164
Available Reserved space: 87
MWI= 93
MD5 =OK
85.07 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.

+1 for the Samsung 840 with TLC. I got more than enough drives to test at the moment so I've to wait out with that one.

[B.A.T]

Here are todays update:

m4
1652.2084 TiB
6783 hours
Avg speed 74.90 MiB/s.
AD 180 to 177
P/E 28429
C3 227775 to 272213
01 83 to 100 (92 to 90)
CE 90
MD5 OK.
Reallocated sectors : 00


Intel X25-E 64GB
1.0 PiB
3273-30=3243 hours
Reallocated sectors : 164 to 175
Available Reserved space: 87 to 86
MWI= 93
MD5 =OK
85.07 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 189

(GiB) 4,666,340
(TiB) 4,556
(PiB) 4.48


(Avg) 297.53 MB/s over the past 1600+ hours

(B1) Wear Leveling Count: 20,186

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

(POH) 4585


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

[Highendtoys]

I was thinking about doing this to a 20nm Micron flash drive I have.

TLC vs. 20nm maybe?