SSD Write Endurance 25nm Vs 34nm

[Christopher]

I haven't decided yet, not sure that I've got one that really needs testing, some are already getting a bit old...
If I can't make up my mind in due time for the 1 year "anniversary" for the thread I'll find some way to randomly select one.

I'm on the lookout for a new #3 drive. There just isn't anything screaming out "test me! test me!".

I'm considering an Intel 330, but I think I want something a little further off the beaten path. Something bizarre perhaps. Or I still have the Mushkin Chronos Deluxe 120GB, but if I run it, I may do so overprovisioned.

I'll leave the Octane in the chart. (until it can be physically checked)

When I finally make it back to the homestead, I'll try and use the OCZ bootable .iso to destructive flash 1.13 again, but I do expect a steaming plate of failure. Until that time, I'm hoping it doesn't crash the endurance rig.

[canthearu]

I've had to turn off my endurance rig for today as it was malfunctioning and I didn't have time to fix it. It will be a quick 10-11 hour retention test!

I think the Intel RST 10.8 drivers are causing a problem, as the Intel 520 has disconnected a couple of times since installing them a couple of days ago. Funny enough, I can still read the SMART from the Intel 520 so the drive certainly isn't dead. I'll try a newer version, but worst comes to worse, I'll revert back to MSAHCI. (but the torqx 2 is 25% faster writing with ioStor, so hopefully I can get it working)

[Christopher]

I've had to turn off my endurance rig for today as it was malfunctioning and I didn't have time to fix it. It will be a quick 10-11 hour retention test!

I think the Intel RST 10.8 drivers are causing a problem, as the Intel 520 has disconnected a couple of times since installing them a couple of days ago. Funny enough, I can still read the SMART from the Intel 520 so the drive certainly isn't dead. I'll try a newer version, but worst comes to worse, I'll revert back to MSAHCI. (but the torqx 2 is 25% faster writing with ioStor, so hopefully I can get it working)

Which RST OROM is your mobo running? I'm sticking with 10.6 for a while.

[canthearu]

Which RST OROM is your mobo running? I'm sticking with 10.6 for a while.

I have never checked ... old P35 board here, so probably running 7.5 or. I've kept an eye on the situation, but since I've never had a problem with my boards so I've never been particularly worried about it.

I can possibly build an updated rom and include a newer OROM (I've done an older nvidia ASUS board some time back)

[Christopher]

Keep us in the loop. I wonder if perhaps the 520's SF heritage is rearing is questionable head. I don't think so, as you would have perhaps experienced the issue earlier.

[canthearu]

It is likely just the intel drivers giving me grief. the msahci driver worked perfectly for 2 months before this problem started happening.

But will keep updated

[Anvil]

I'm pretty sure I have been using 10.8 since it was released, that is on the Z68 used for Endurance testing.
I'll double check, anyways I've been using the Intel drivers all along as I always install in raid-mode.

edit:
It is 10.8.0.1003, RAID option ROM is 10.6.0.1091

The driver is dated mid November 2011.

[canthearu]

Well, I've gone back to msahci drivers for now.

My endurance system motherboard is very old now, so throwing the latest drivers on it doesn't always work out.

[Anvil]

Could be a bad combo of MB and driver although I'm pretty sure that the drivers should be backwards compatible.

Are the drives back up and running?

[canthearu]

Yep, drives came through 10 hours (turned off) without a problem. Endurance testing restarted

I'll let them go for a week. If I have no more problems, I'll be happy that it isn't a drive problem and I'll try the older driver's of Asus website.

[B.A.T]

I know this is an Plextor but with that size it looks interesting

Link

They already got it in stock at several Norwegian webshops.

[johnw]

The 64GB Plextor M3 has been available in the US (newegg) for a while (1-2 months?). I think I even posted about it on XS when it first showed up.

[B.A.T]

I've seem to have missed that one.

[B.A.T]

Todays update:

Intel X25-M G1 80GB
519.80 TiB
22409 hours
Reallocated sectors : 93 to 97
Available Reserved space: 94
MWI=205 to 199
MD5 =OK
42.05 MiB/s on avg


m4
718.0030 TiB
3074 hours
Avg speed 73.10 MiB/s.
AD gone from 201 to 198.
P/E 12434.
MD5 OK.
Reallocated sectors : 00


Only 50 TiB until the last m4 record is breaken

[Christopher]

Samsung 830 256 Day 31

719,649.19GiB
702.78TiB
289.97 MB/s Average

3142 Wear Leveling Count
MWI 13

0/0 Erase/Program Fail

OCZ Vertex Turbo 64 Day 48

359,903.87GiB
351.46TiB
84.55 MB/s Avg

6432 Avg PE count
-100 MWI

0/0/0 Read/Program/Erase Fail

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

[canthearu]

Decided that I'm too lazy to reinstall windows on my desktop, it is working fine and touching it is bad ... so the vertex 4 goes into the endurance rig. Besides, the octane drive failing early doesn't give me the most confidence in this drive.

I've updated the OROM to 8.9 on my P5K-E/Wifi-AP motherboard, using the BIOS links provided by OCZ. Hopefully that will resolve the stability difficulties I have been having.

Just copying static data to drive now . . . . stupid wireless network.

Endurance system stats (this is for all drives)

Intel Pentium E5400 at 2.7ghz (stock)
Asus P5K-E/Wifi-AP motherboard (modified BIOS, 8.9 OROM)
4gig RAM
Windows 7 x64

Will post again with performance figures and drive settings once this copy is done.

[canthearu]

Ok, drive information:

Compression: 46% (picture will show 0%, I changed it right after)
Static Data: 58.6 GiB

Some benchies:


CrystaldiskInfo ... Smart is working:


Endurance (quick compared to my intel 520):

[canthearu]

Patriot Torqx-2 64GB - Day 62

Drive hours: 1,501
ASU GiB written: 99,295.62 GiB (96.9685 TiB)
Avg MB/s: 25.82 MB/s
MD5: OK

Bad blocks: 2 grown, 83 factory
Wear cycle counter: 6232 average erase cycles, 9553 maximum erase cycles.
Total ECC Error Count: 1

Intel 520 60GB - Day 54

Drive hours: 1,299
ASU GiB written: 403,933.5 GiB (394.46 TiB)
Avg MB/s: 92.15 MB/s
MD5: OK

Host GB written (F1): 406,743.78 GiB (397.21 TiB, 13015801 raw)
NAND writes (F9): 288,013 GiB (281.26 TiB)

Reallocated sectors (05): 0
Failure count (AB, AC): 0 program, 0 erase
Raw Error Rate (8B): 120 normalised
Avaliable Reserved Space (AA): 100 normalised
Media Wearout Indicator (E9): 7

Vertex 4 128GB - Day 1

Drive hours: 24
ASU GiB written: 1,033.52 GiB (1.01 TiB)
Avg MB/s: 160.12 MB/s
MD5: OK

Sector GB written (E8): 1,766.08 GiB (1.72 TiB, 3703733049 raw)

Reallocated sectors (05): 0
Raw Read Error Rate (01): 7
Remaining Life (E9): 100

[B.A.T]

Todays update:

Intel X25-M G1 80GB
523.16 TiB
22431 hours
Reallocated sectors : 97 to 101
Available Reserved space: 94 to 93
MWI=199 to 193
MD5 =OK
43.34 MiB/s on avg


m4
723.6908 TiB
3097 hours
Avg speed 73.08 MiB/s.
AD gone from 198 to 195.
P/E 12529.
MD5 OK.
Reallocated sectors : 00

[Christopher]

Let's see how the Vertex4 does as a contrast to the Octane.

EDIT
It should [The Vertex4] should perform like the Octane but with better ECC. The Avg MB/s are very similar.

[Anvil]

Endurance_cr_20120419.png

I'll update post 1 as well with the new drive.

@canthearu
I've put IMFT as NAND maker, did you check if it was Intel or Micron?
I've also put 512MB but I'm not sure if the 128GB drive used 256MB or 512MB?

--

Kingston SSDNow 40GB (X25-V)

885.02TB Host writes
Reallocated sectors : 05 41
Available Reserved Space : E8 99
POH 7946
MD5 OK

33.27MiB/s on avg (~217 hours)

[canthearu]

@Anvil

I havn't cracked it open. I don't want to void the warrenty if I get a particularly early death. Should be IMFT.
Cache is 512meg (actually 1gig, but only 512meg active, as per anandtech review)

Controller is: Indilinx Everest 2 (yes, it is based on the new marvell controller, but given indilinx firmware is quite different, it shouldn't be lumped in with other marvell drives for now)

[canthearu]

Patriot Torqx-2 64GB - Day 63

Drive hours: 1,524
ASU GiB written: 101,350.75 GiB (98.97 TiB)
Avg MB/s: 26.16 MB/s
MD5: OK

Bad blocks: 2 grown, 83 factory
Wear cycle counter: 6361 average erase cycles, 9749 maximum erase cycles.
Total ECC Error Count: 1

Intel 520 60GB - Day 55

Drive hours: 1,322
ASU GiB written: 411,510.79 GiB (401.87 TiB)
Avg MB/s: 90.57 MB/s
MD5: OK

Host GB written (F1): 414,391.28 GiB (404.68 TiB, 13260521 raw)
NAND writes (F9): 293,423 GiB (286.55 TiB)

Reallocated sectors (05): 0
Failure count (AB, AC): 0 program, 0 erase
Raw Error Rate (8B): 109 normalised
Avaliable Reserved Space (AA): 100 normalised
Media Wearout Indicator (E9): 5

Vertex 4 128GB - Day 2

Drive hours: 53
ASU GiB written: 13,705.07 GiB (13.38 TiB)
Avg MB/s: 142.79 MB/s
MD5: OK

Sector GB written (E8): 14,505.28 GiB (14.17 TiB, 30419775523 raw)

Reallocated sectors (05): 0
Raw Read Error Rate (01): 6
Remaining Life (E9): 99

Notes:

* Still having problems with Intel rst driver 10.8. Trying 8.9 from asus's website now .... though it is somewhat slower.
* 2 days until the Intel 520 reaches MWI 1
* The MWI works on the Vertex 4

[devsk]

Looks like samsung is going to break 1PiB before anything else. 290MB/s avg write speed is 2-4 times faster than the next best things.

[Christopher]

Looks like samsung is going to break 1PiB before anything else. 290MB/s avg write speed is 2-4 times faster than the next best things.

maybe, but its use of PE cycles is agonizingly slow. Even though it's 3K PE rated flash, 177 is still at 8/8/3342 -- so I don't get it. I think it [MWI] may be based on 4K PE cycles.





Samsung 830 256 Day 33

766,392.54GiB
748.43TiB
291.00 MB/s Average

3342Wear Leveling Count
MWI 8

0/0 Erase/Program Fail

OCZ Vertex Turbo 64 Day 50

373,049.81GiB
364.29TiB
84.55 MB/s Avg

6666 Avg PE count
-100 MWI

0/0/0 Read/Program/Erase Fail
to
------------------------------------------------------
Both drives continue to get ever so slightly faster over time. The 830 was averaging in the low/mid 280s weeks ago and is now at 291MB/s. The VT continues to gain a 10th of a MB/s every so often.