SSD Write Endurance 25nm Vs 34nm

[Anvil]

Kingston SSDNow 40GB (X25-V)

442.42TB Host writes
Reallocated sectors : 12

MD5 OK

33.13MiB/s on avg (~217 hours)

--

Corsair Force 3 120GB

01 90/50 (Raw read error rate)
05 2 (Retired Block count)
B1 73 (Wear range delta)
E6 100 (Life curve status)
E7 25 (SSD Life left)
E9 285083 (Raw writes)
F1 379474 (Host writes)

MD5 OK

103.12MiB/s on avg (~217 hours) 9 days 1 hour, 76.75TiB written so far during this session.

power on hours : 1123

B1 (WRD) is down again, from 74 to 73.

[Christopher]

Mushkin Chronos Deluxe 60 Update, Day 32
3.3.2FW

05 4
Retired Block Count

B1 11
Wear Range Delta

F1 399688
Host Writes

E9 308340
NAND Writes

E6 100
Life Curve

E7 10
Life Left

Average 128.14/s Avg
Intel RST drivers, Celeron G530 Biostar TH67+

947 Hours Work (24hrs since the last update)
Time 1 month 9 days 11 hours

12 GiB Minimum Free Space

176.8 hours uptime

[overdoze]

Great Thread I thank you all that participate in testing

[Christopher]

Milestone reached

310TB/402TB

Even if it is slightly arbitrary' its still a milestone.

[bulanula]

Is the Samsung official dead now ? No more recovery attempts ?

[Anvil]

Kingston SSDNow 40GB (X25-V)

443.59TB Host writes
Reallocated sectors : 12

MD5 OK

33.11MiB/s on avg (~227 hours)

--

Corsair Force 3 120GB

01 90/50 (Raw read error rate)
05 2 (Retired Block count)
B1 72 (Wear range delta)
E6 100 (Life curve status)
E7 24 (SSD Life left)
E9 287906 (Raw writes)
F1 383230 (Host writes)

MD5 OK

103.10MiB/s on avg (~227 hours) >80TiB Host writes during this session.

power on hours : 1133

B1 (WRD) is down again, from 73 to 72.

[Anvil]

Regarding the charts.

We'll give it this week and if there is no word we'll have to find someone that can temporarily do the charts.

[Ao1]

It would be good to have a link to the excel charts in the first post that are refreshed every time the charts are updated. That way anyone can look and play with the data and if someone drops out it’s not a matter of starting from scratch.

There is no point doing this work if the data collected is not properly recorded and analysed. There are a lot of variables that don’t seem to be recorded. What compression factor has been used? What is the ratio of static data to free space? Events such as the deletion of static data are not being captured. Nand/ P/E cycle specs are not clear. Controller type/ version. Is the drive being tested an OS drive. What OS etc. etc.

I would like to suggest that each person testing is responsible to collate data for their drive in a pre agreed standard excel format and that one person is responsible for collating data from all the drives.

[bluestang]

M225->Vertex Turbo 64GB Update:

649.68 TiB (714.34 TB) total
1716.03 hours
12931 Raw Wear
123.19 MB/s avg for the last 27.31 hours (on W7 x64)
MD5 OK
C4-Erase Failure Block Count (Realloc Sectors) at 11.
(1=Bnk 6/Blk 2406; 2=Bnk 3/Blk 3925; 3=Bnk 0/Blk 1766; 4=Bnk 0/Blk 829; 5=Bnk 4/Blk 3191; 6=Bnk 7/Blk 937; 7=Bnk 7/Blk 1980; 8=Bnk 7/Blk 442; 9=Bnk 7/Blk 700; 10=Bnk 2/Blk 1066; 11=Bnk 7/Blck 85)

[EDIT] P.S. -- Are they ever going to fix the picture attachment problem or what?

[Christopher]

Ao1,

Most ofthe data you speak of was mostly represented, at least until attachments broke.

WRD looks to be rising again after bottoming out at 9.Currently 12.

[Anvil]

Ao1

I agree for the most part
It will make it easier for the one in charge of doing charts that input is given in a standard/common format so we'll have to do something about that.

As for what's recorded or not
The Endurance app records every byte written and what compression was/is used and as long as the log file is present one can easily find
- compression level used for each loop
- number of files and GB written for each loop

What is written and/or deleted outside the Endurance app is unknown, I'm not sure that it really matters that much as long as it's done to keep the drive up to date (as in fw updates and/or secure erasing due to issues) and not for other purposes.

Deletion of static files might be useful to record and is recorded in my case. (only applicable for the SF drive due to fw issues)

The exact P/E spec is known for the drives that have been opened, for the other drives MWI points to the expected P/E count for the drives. (there are a few drives that have not been opened for inspection yet, including mine )
I see no point in opening my drives yet, I might do that when a data retention test is due for the drives. (BAT is having one now)

So, there is very little that's unknown/unaccounted for but we should of course agree on a standard for reporting and reporting issues such as cleaning, changing-/updating static files.
The drive (F40) used by SynbiosVyse is the one drive that I'm a bit uncertain of and I expect the log file is gone as the drive stopped working.

[bluestang]

This is how I'm keeping track of mine...

[Christopher]

Anvil,

is there any way to add a feature to ASU that will store the log on another drive? or a server?

[Anvil]

@bluestang

Great work, I keep screenshots for every reporting but using a spreadsheet is what we all should do

@Christopher
Not currently, I copy my "Endurance" log files to a dedicated folder on the boot drive for each drive used in the test.
I have thought about finding some other way to store the file but it requires to be stored in a subfolder of the Endurance app so the only thing that could change this is allowing the app to be run from a separate drive.
It might be an idea to "export" the data for storage on a "server", I'll have to think about that.

I could make it post updates to a webserver at some interval, a lot of options...

[zads]

Looks expensive
What's the workload?
...

Expensive (especially considering the time consumption) is right, 35x 128GB eMLC drives..
Workload is just a general/typical smattering of accesses- block sizes, rand/seq, comp/incomp data.

In December I'm doing SSD radiation bombardment testing..

[bluestang]

@bluestang

Great work, I keep screenshots for every reporting but using a spreadsheet is what we all should do

I have both, a spreadsheet and screenshots. I like overkill

[Anvil]

Kingston SSDNow 40GB (X25-V)

444.88TB Host writes
Reallocated sectors : 12

MD5 OK

37.74MiB/s on avg (~4 hours)

--

Corsair Force 3 120GB

01 88/50 (Raw read error rate)
05 2 (Retired Block count)
B1 71 (Wear range delta)
E6 100 (Life curve status)
E7 23 (SSD Life left)
E9 291427 (Raw writes)
F1 387915 (Host writes)

MD5 OK

103.09MiB/s on avg (>240 hours) ~85TiB Host writes during this session.

power on hours : 1146

B1 (WRD) is down again, from 72 to 71.

[Christopher]

Mushkin Chronos Deluxe 60 Update, Day 32
3.3.2FW


05 4
Retired Block Count

B1 10
Wear Range Delta

F1 413311
Host Writes

E9 318862
NAND Writes

E6 100
Life Curve

E7 10
Life Left

Average 128.44MB/s Avg
Intel RST drivers, Celeron G530 Biostar TH67+

977 Hours Work
Time 1 month 10 days 17 hours

12 GiB Minimum Free Space

207hrs uptime

[SteveRo]

Zads - considering SSD space qualification?

[Anvil]

Kingston SSDNow 40GB (X25-V)

446.19TB Host writes
Reallocated sectors : 12

MD5 OK

35.09MiB/s on avg (~16 hours)

--

Corsair Force 3 120GB

01 94/50 (Raw read error rate)
05 2 (Retired Block count)
B1 70 (Wear range delta)
E6 100 (Life curve status)
E7 22 (SSD Life left)
E9 294481 (Raw writes)
F1 391977 (Host writes)

MD5 OK

103.09MiB/s on avg (>251 hours) ~89TiB Host writes during this session.

power on hours : 1157

B1 (WRD) is down again, from 71 to 70.

[bluestang]

M225->Vertex Turbo 64GB Update:

656.57 TiB (721.91 TB) total
1731.92 hours
13056 Raw Wear
114.63 MB/s avg for the last 17.40 hours (on W7 x64)...switched to Beta9 and lost some speed
MD5 OK
C4-Erase Failure Block Count (Realloc Sectors) at 11.
(1=Bnk 6/Blk 2406; 2=Bnk 3/Blk 3925; 3=Bnk 0/Blk 1766; 4=Bnk 0/Blk 829; 5=Bnk 4/Blk 3191; 6=Bnk 7/Blk 937; 7=Bnk 7/Blk 1980; 8=Bnk 7/Blk 442; 9=Bnk 7/Blk 700; 10=Bnk 2/Blk 1066; 11=Bnk 7/Blck 85)

[devsk]

M225->Vertex Turbo 64GB Update:

656.57 TiB (721.91 TB) total
1731.92 hours
13056 Raw Wear
114.63 MB/s avg for the last 17.40 hours (on W7 x64)...switched to Beta9 and lost some speed
MD5 OK
C4-Erase Failure Block Count (Realloc Sectors) at 11.
(1=Bnk 6/Blk 2406; 2=Bnk 3/Blk 3925; 3=Bnk 0/Blk 1766; 4=Bnk 0/Blk 829; 5=Bnk 4/Blk 3191; 6=Bnk 7/Blk 937; 7=Bnk 7/Blk 1980; 8=Bnk 7/Blk 442; 9=Bnk 7/Blk 700; 10=Bnk 2/Blk 1066; 11=Bnk 7/Blck 85)

If my vertex 1 lives to see one tenth of those writes, I would be very happy! I have written about 1TB on 120G in one year with normal usage.

[Christopher]

Wow, we should start endurance testing while being bombarded with lethal amounts of radiation.

Actually, a month ago if you had told me that I could fix my SandForce problems with lethal quantities of radiation, I'd have tried it.

[Anvil]

Kingston SSDNow 40GB (X25-V)

447.56TB Host writes
Reallocated sectors : 12

MD5 OK

34.07MiB/s on avg (~28 hours)

--

Corsair Force 3 120GB

01 95/50 (Raw read error rate)
05 2 (Retired Block count)
B1 69 (Wear range delta)
E6 100 (Life curve status)
E7 21 (SSD Life left)
E9 297766 (Raw writes)
F1 396350 (Host writes)

MD5 OK

103.09MiB/s on avg (>263 hours) ~93TiB Host writes during this session.

power on hours : 1170

B1 (WRD) is down again, from 70 to 69.

@johnw

Did you mention something about testing the Samsung 830?

[Christopher]

Mushkin Chronos Deluxe 60 Update, Day 43
3.3.2FW

05 4
Retired Block Count

B1 10
Wear Range Delta

F1 421680
Host Writes

E9 325324
NAND Writes

E6 100
Life Curve

E7 10
Life Left

Average 128.35MB/s Avg
Intel RST drivers, Celeron G530 Biostar TH67+

995 Hours Work
Time 1 month 11 days 11 hours

12 GiB Minimum Free Space

229.9 hours uptime