SSD Write Endurance 25nm Vs 34nm

[bulanula]

Too bad there is no C300 in this test anymore. I was really curious as to how it would perform. I think it would have thrashed the M4 and maybe even the Turbo drive. Good to hear all is OK.

[Anvil]

@Christopher

I'll dig up what I found when testing my 830's but you got F1 right, not sure if/what else I found.
Great to see that it's as speedy as expected

The Samsung's really are something special!

--

Kingston SSDNow 40GB (X25-V)

529.33TB Host writes
Reallocated sectors : 05 14
Available Reserved Space : E8 99

MD5 OK

33.28MiB/s on avg (~61 hours)

--

Corsair Force 3 120GB

01 92/50 (Raw read error rate)
05 2 (Retired Block count)
B1 49 (Wear range delta)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 482647 (Raw writes) ->471TiB
F1 642402 (Host writes) ->627TiB

MD5 OK

106.34MiB/s on avg (~61 hours)

power on hours : 1841

[Anvil]

M225->Vertex Turbo 64GB Update:

Soooo, I started the test back up this morning after letting it sit idle over the weekend...It ran for 104 loops/3.63 hrs at 117.61 MB/s and C5 jumped from 7 to 15 in that time. MD5 in ASU still showed OK. Stopped test as I needed to reboot after an update and afterwards drive was unresponsive, but still showed up in BIOS and OS. Mulitiple attempts of unplugging/powering off did not help this time. So, I was forced to do a full format and lost ~2 TiB of logging from ASU (I have ASU data from Friday though).
...
868.82 TiB (955.28 TB) total.
~2224 hrs (Torture), 3257 hrs (Power-On).
...

As long as SMART is working and you've got the total there really is no problem, let's hope that it doesn't happen again

[bluestang]

As long as SMART is working and you've got the total there really is no problem, let's hope that it doesn't happen again

Just logged in to my work PC and checked on it... 958 TB and going!
Bad block counts haven't moved since last update...Let's hope I didn't just jinx it!!

[Christopher]

Just about 2000 GiB in 5 hrs isn't too slow for a 64GB drive without deduplication, especially one built on 25nm NAND.

The only annoying thing about the Samsung is the Magician software suite does everything -- except the arguably more important stuff, like giving you comprehensible SMART info.


-------------------------------------------------------------------------------------------------------
I have another drive in the mail. I'd love to tell you what it is, but I actually have no idea myself. It's a Patriot of some kind.

I might have just violated rule number #273: Don't buy a J.Micron controlled SSD.

If it's not the worst drive in the world, we could have another addition.

EDIT...
I think it's a Phison PS3016-S3

[Anvil]

The next one I'll test will most likely be one of the drives already tested but with extra over-provisioning.

There's not much to find on the NAND Samsung has put into the new series, could be 27nm (or 22nm or 24nm), I can't find any sources but the processor is triple core ARM9, model# S4LJ204X01

[Christopher]

I think I found out why the 830 has such great sequential writes.... it's totally lazy on the garbage collection side of things.

My speeds were continuously excellent until I fixed the random writes (I added an extra second to ensure the 100MB randoms were always written.) After I did that, speeds plunged down into the 78MB/s range after after a dozen or so loops. I'm trying to tweak some settings to preserve speed and random writes too.

The Samsung user initiated GC/optimization can be scheduled like Intel's Toolbox, but takes 4 to 5 minutes to run. Maybe I can find a series of settings that can keep speeds more constant like the Mushkin's were.

-----------------
The first of the 24nm Toshiba toggle NAND products are out (the Plextor 3 series). Given the relationship between Toshiba and Samsung, they may be using similar stuff 24nm processes. I've read people who say that the 830's nand is 27nm, 25nm, and 24nm. I don't really care... I like it.

[minpayne]

A faulty M4 512GB I'm trying to RMA:



Only Anvil's Storage Utilities Endurance testing and WinHex fill disk sectors can reliably force the disk to go offline like this and reproduce the problem. I know that something is definitely wrong with some nand flash, as the 05 SMART says 413k+ reallocated sector count.

[bluestang]

M225->Vertex Turbo 64GB Update:

Drive dropped out last night, looked like during file delete. Came in this morning to see ASU write errors. A quick unplug and re-plug of the drive and back online. C5 jumped again, up by 6...Don't know how much more of this the M225 will take.

873.85 TiB (960.80 TB) total.
~2232 hrs (Torture), 3273 hrs (Power-On).
16992 Raw Wear.
??? MB/s avg for the last ??? hours (on W7 x64).
MD5 every 20 loops on 1.59GB file = OK.
C4-Erase Failure Block Count (Realloc Sectors) at 24.
C5-Read Failure Block Count (uncorrectable bit errors) from 15 to 21. again
1=Bnk 6/Blk 2406 - Erase Failure C4
2=Bnk 3/Blk 3925 - Erase Failure C4
3=Bnk 0/Blk 1766 - Erase Failure C4
4=Bnk 0/Blk 829 - Erase Failure C4
5=Bnk 4/Blk 3191 - Erase Failure C4
6=Bnk 7/Blk 937 - Erase Failure C4
7=Bnk 7/Blk 1980 - Erase Failure C4
8=Bnk 7/Blk 442 - Erase Failure C4
9=Bnk 7/Blk 700 - Erase Failure C4
10=Bnk 2/Blk 1066 - Erase Failure C4
11=Bnk 7/Blck 85 - Erase Failure C4
12=Bnk 4/Blk 3192 - Erase Failure C4
13=Bnk 7/Blk 280 - Erase Failure C4
14=Bnk 3/Blk 2375 - Erase Failure C4
15=Bnk 7/Blk 768 - Erase Failure C4
16=Bnk 7/Blk 765 - Erase Failure C4
17=Bnk 7/Blk 182 - Erase Failure C4
18=Bnk 5/Blk 939 - Read Failure C5
19=Bnk 5/Blk 1115 - Read Failure C5
20=Bnk 5/Blk 1011 - Read Failure C5
21=Bnk 7/Blk 3549 - Read Failure C5
22=Bnk 7/Blk 3556 - Read Failure C5
23=Bnk 4/Blk 1961 - Erase Failure C4
24=Bnk 7/Blk 1862 - Erase Failure C4
25=Bnk 7/Blk 111 - Erase Failure C4
26=Bnk 5/Blk 902 - Read Failure C5
27=Bnk 7/Blk 560 - Erase Failure C4
28=Bnk 7/Blk 827 - Erase Failure C4
29=Bnk 6/Blk 482 - Erase Failure C4
30=Bnk 5/Blk 167 - Read Failure C5
31=Bnk 7/Blk 1771 - Erase Failure C4
32=Bnk 5/Blk 956 - Read Failure C5
33=Bnk 5/Blk 1242 - Read Failure C5
34=Bnk 5/Blk 1461 - Read Failure C5
35=Bnk 5/Blk 1800 - Read Failure C5
36=Bnk 5/Blk 1030 - Read Failure C5
37=Bnk 5/Blk 1074 - Read Failure C5
38=Bnk 5/Blk 1128 - Read Failure C5
39=Bnk 5/Blk 1427 - Read Failure C5
40=Bnk 0/Blk 1674 - Read Failure C5
41=Bnk 5/Blk 756 - Read Failure C5
42=Bnk 5/Blk 788 - Read Failure C5
43=Bnk 5/Blk 909 - Read Failure C5
44=Bnk 5/Blk 407 - Read Failure C5
45=Bnk 5/Blk 1294 - Read Failure C5

So a total (B8+C3+C4+C5) of 55 Bad Blocks

Edit: Drive has slowed down some also to around 112 MB/s.

[Anvil]

Lets see if it can manage another 40TB, should go past 1PB in less than a week if it doesn't slow down too much.

[bluestang]

Lets see if it can manage another 40TB, should go past 1PB in less than a week if it doesn't slow down too much.

I'm giving it my all...and so is the M225!

[Computurd]

wow unbelievable, getting to that magic number! 1 PB that is just crazy...

[Christopher]

I love that M225 "Turbo".

I can't help but feel like it's on it's last legs though... please, prove me wrong.

[bluestang]

And I still have another M225 sitting in a notebook doing nothing

Who knows, maybe I'll flash that to FW 1.7 (I heard there is better WA/endurance with it)

[Computurd]

is there a 1.7 for the indys? is that arrowana or ?? dont know i lost track on that whole deal. I actually fired up my vertexes yesterday, was thinking firmwares...

[RyderOCZ]

1.7 is not Arowana.

I thought this was a real Turbo, you mean you counterfeited the drive?

[bluestang]

I thought this was a real Turbo, you mean you counterfeited the drive? :P

[Christopher]

1.7 is not Arowana.

I thought this was a real Turbo, you mean you counterfeited the drive?

The saga of the M225 Turbo starts in the Turbo thread in the Indilinx forums.

We're all so very proud of Bluestang's initiative.


I bought a new Vertex Turbo 128 (there were a couple floating around for a short time a couple months ago). It's great, but I think the WA with smaller writes is pretty high.

[Christopher]

Samsung 830 64GB Update Day 1

GiB written :
9116.33

Avg MB/s :
108.25

Time Elapsed:
24hrs



I'm not really sure what else to put in the updates at the moment.

[B.A.T]

Todays update:
Kingston V+100
280.9583 TiB
1259 hours
Avg speed 28.83 MiB/s
AD still 1.
168= 1 (SATA PHY Error Count)
P/E?
MD5 OK.
Reallocated sectors : 00


The Kingston has disconnected 3 times during the 3 last nights... Something is not right with this one. Same ting with the low write speed.

Intel X25-M G1 80GB
100,7683 TiB
19410 hours
Reallocated sectors : 00
MWI=71 to 70
MD5 =OK
51.88 MiB/s on avg


m4
6.6668 TiB
24 hours
Avg speed 80.19 MiB/s.
AD gone from 100 to 97.
P/E 110.
MD5 OK.
Reallocated sectors : 00

[bluestang]

M225->Vertex Turbo 64GB Update:

876.17 TiB (963.36 TB) total.
~2239 hrs (Torture), 3280 hrs (Power-On).
17036 Raw Wear.
111.00 MB/s avg for the last 6.05 hours (on W7 x64).
MD5 every 20 loops on 1.59GB file = OK.
C4-Erase Failure Block Count (Realloc Sectors) at 24.
C5-Read Failure Block Count (uncorrectable bit errors) from 21 to 24.
1=Bnk 6/Blk 2406 - Erase Failure C4
2=Bnk 3/Blk 3925 - Erase Failure C4
3=Bnk 0/Blk 1766 - Erase Failure C4
4=Bnk 0/Blk 829 - Erase Failure C4
5=Bnk 4/Blk 3191 - Erase Failure C4
6=Bnk 7/Blk 937 - Erase Failure C4
7=Bnk 7/Blk 1980 - Erase Failure C4
8=Bnk 7/Blk 442 - Erase Failure C4
9=Bnk 7/Blk 700 - Erase Failure C4
10=Bnk 2/Blk 1066 - Erase Failure C4
11=Bnk 7/Blck 85 - Erase Failure C4
12=Bnk 4/Blk 3192 - Erase Failure C4
13=Bnk 7/Blk 280 - Erase Failure C4
14=Bnk 3/Blk 2375 - Erase Failure C4
15=Bnk 7/Blk 768 - Erase Failure C4
16=Bnk 7/Blk 765 - Erase Failure C4
17=Bnk 7/Blk 182 - Erase Failure C4
18=Bnk 5/Blk 939 - Read Failure C5
19=Bnk 5/Blk 1115 - Read Failure C5
20=Bnk 5/Blk 1011 - Read Failure C5
21=Bnk 7/Blk 3549 - Read Failure C5
22=Bnk 7/Blk 3556 - Read Failure C5
23=Bnk 4/Blk 1961 - Erase Failure C4
24=Bnk 7/Blk 1862 - Erase Failure C4
25=Bnk 7/Blk 111 - Erase Failure C4
26=Bnk 5/Blk 902 - Read Failure C5
27=Bnk 7/Blk 560 - Erase Failure C4
28=Bnk 7/Blk 827 - Erase Failure C4
29=Bnk 6/Blk 482 - Erase Failure C4
30=Bnk 5/Blk 167 - Read Failure C5
31=Bnk 7/Blk 1771 - Erase Failure C4
32=Bnk 5/Blk 956 - Read Failure C5
33=Bnk 5/Blk 1242 - Read Failure C5
34=Bnk 5/Blk 1461 - Read Failure C5
35=Bnk 5/Blk 1800 - Read Failure C5
36=Bnk 5/Blk 1030 - Read Failure C5
37=Bnk 5/Blk 1074 - Read Failure C5
38=Bnk 5/Blk 1128 - Read Failure C5
39=Bnk 5/Blk 1427 - Read Failure C5
40=Bnk 0/Blk 1674 - Read Failure C5
41=Bnk 5/Blk 756 - Read Failure C5
42=Bnk 5/Blk 788 - Read Failure C5
43=Bnk 5/Blk 909 - Read Failure C5
44=Bnk 5/Blk 407 - Read Failure C5
45=Bnk 5/Blk 1294 - Read Failure C5
46=Bnk 5/Blk 562 - Read Failure C5
47=Bnk 5/Blk 256 - Read Failure C5
48=Bnk 5/Blk 943 - Read Failure C5

So a total (B8+C3+C4+C5) of 58 Bad Blocks.

Stopping test for the night, will pick up tomorrow morning when I can babysit my little M225.

[Anvil]

I'm catching up on the summary and I just noticed that the attachments stopped working October the 27th, it's been over 1 month, we deserve an explanation!

[B.A.T]

Shouldn't we put in TRIM and fw ver?

[Anvil]

We could but there are just 2 drives that have been running w/o TRIM and one of them lacked TRIM due to running off XP so it's really just the X25-M G1 that lacks TRIM.
FW would be OK, although I (and Christopher) have changed fw several times during the test so we'd have to list all fw's used.

[Christopher]

I listed the FW version for the Mushkin in the updates after I switched from 3.2.0

The first month of Chronos testing was done with 320fw. Then I went to 3.3.0 for one day, then 3.3.2 was released, so I used it for the remainder.