SSD Write Endurance 25nm Vs 34nm

[johnw]

201.525 TiB, 555 hours, sa177: 1/1/16532

211.627 TiB, 582 hours, sa177: 1/1/17433

Average speed reported by Anvil's app has been steady at about 112MB/s.

The other two unknown SMART attributes, 178 and 235, are still at 72/72/276 and 99/99/2, just as they were when the SSD was fresh out of the box.

[B.A.T]

And then I back again:

130.9436 TiB
442 hours
Avg speed 90.96 MiB/s.
AD gone from 26 to 24.
P/E 2283.
MD5 OK.

[RejZoR]

I'm really wondering how far will these go before they are degraded beyond usable. And how will that look like. I wish these test files were usable files which could be later inspected if they are still usable when SSD is starting to show severe symptoms od degradation. Because to date, i don't think anyone has seen a drive that was degraded that much. Anyway, very interesting test, too bad there aren't any SandForce 22xx series drives in it (like Corsair Force 3 and OCZ Vertex 3)...

[Anvil]

179.08TB Host writes
WMI 2
Reallocated sectors : 6

MD5 OK, 33.1MiB/s on avg. (28 hours)

[Vapor]

C300 Update

91.914TiB, 69 MWI, 1551 raw wear indicator, 1021590229 0xF8 value, 61.85MiB/sec, MD5 OK.

New: 2048 reallocated sectors; 1 reallocation event count; 1 grown failing block count; 1 erase fail count.

[Anvil]

Just thought I had to post this milestone

2011-07-23-21-09_180TIB.PNG

[Computurd]

excellent!!

been a long time coming, that kingston is a champ

[johnw]

211.627 TiB, 582 hours, sa177: 1/1/17433

219.480 TiB, 603 hours, sa177: 1/1/18053

Average speed reported by Anvil's app has been steady at about 112MB/s.

The other two unknown SMART attributes, 178 and 235, are still at 72/72/276 and 99/99/2, just as they were when the SSD was fresh out of the box.

10*365.25*64e9/1024^4 = 212.603TiB

Past the point where the writes are equivalent to writing the entire 64GB capacity of the SSD every day for 10 years.

[Computurd]

Past the point where the writes are equivalent to writing the entire 64GB capacity of the SSD every day for 10 years.

well thats a stat you can sink your teeth into! very impressive! unbelievable actually

glad to see this moving along...getting a tad boring with no failures and this has been going on for two months or so....

(dont get me wrong, its a good type of boring )

[sergiu]

C300 Update

91.914TiB, 69 MWI, 1551 raw wear indicator, 1021590229 0xF8 value, 61.85MiB/sec, MD5 OK.

New: 2048 reallocated sectors; 1 reallocation event count; 1 grown failing block count; 1 erase fail count.

Finally a little action... yet I'm not sure how I can interpret the data:
- when we have a reallocated sector, it means that block can still be erased but a certain page cannot be programmed?
- when we have a reallocated block it means that erase failed?
If this is correct, it means that you have a failed block and 1920 failed pages spread to at least 15 blocks (this assuming 4KB pages/128 pages per block). I would expect a rapid failure also for the other blocks that contain failed pages, maybe in next TB written.

[Vapor]

Samsung 470 pseudo-passed the Intel 320 in writes, damn it's fast.

I'm pretty sure the 320 is still ahead, but it's been a few days since the last update on the 320 40GB and it's capable of racking up more than 2.5TiB in that time frame.

Very impressive from the 470 considering it appears to have a WA of ~5.1x

(very impressive of all the drives so far, really, with a partial exception of the V2 40GB's LTT)

[Vapor]

Finally a little action... yet I'm not sure how I can interpret the data:
- when we have a reallocated sector, it means that block can still be erased but a certain page cannot be programmed?
- when we have a reallocated block it means that erase failed?
If this is correct, it means that you have a failed block and 1920 failed pages spread to at least 15 blocks (this assuming 4KB pages/128 pages per block). I would expect a rapid failure also for the other blocks that contain failed pages, maybe in next TB written.

I think Crucial/Micron considers an LBA sector (512B) as their 'sector' (as compared to Intel doing something else, it seems). So 2048 reallocated sectors is just 1MB. I don't know the specs of the NAND in the C300, but that's probably only 1 or 2 blocks--probably 1 based on "1 grown failing block count".

[sergiu]

It makes sense, so probably we could consider this one block as one "Intel sector".

[johnw]

I think Crucial/Micron considers an LBA sector (512B) as their 'sector' (as compared to Intel doing something else, it seems). So 2048 reallocated sectors is just 1MB. I don't know the specs of the NAND in the C300, but that's probably only 1 or 2 blocks--probably 1 based on "1 grown failing block count".

My guess is that they do not want to deal with erase blocks with any bad pages. So if any piece of an erase block fails, I guess they remap the entire erase block.

[bulanula]

Nice to see these SSDs lasting as much as they are doing now !

[B.A.T]

M4 update:

143.1145 TiB
481 hours
Avg speed 89.77 MiB/s.
AD gone from 24 to 17.
P/E 2493.
MD5 OK.

[Anvil]

183.23TB Host writes
MWI 1
Reallocated sectors : 6

MD5 OK.

[johnw]

219.480 TiB, 603 hours, sa177: 1/1/18053

230.010 TiB, 631 hours, sa177: 1/1/18925

Average speed reported by Anvil's app has been steady at about 112MB/s.

The other two unknown SMART attributes, 178 and 235, are still at 72/72/276 and 99/99/2, just as they were when the SSD was fresh out of the box.

[Vapor]

C300 Update

98.07 TiB, 67MWI, 1656 raw wear, MD5 OK, reallocated event still 1 and 2048 sectors. Switched to an updated version of Anvil's app, average speed over the past 8 hours is down to 60.5MiB/sec from 61.85ish.

[nn_step]

First update:


OCZ Agility 3 AGT3-25SAT3-120G - a couple minor sector failures on one of the drives.
Corsair Force CSSD-F120GB2-BRKT - Nothing yet to report
OCZ Vertex 2 OCZSSD2-2VTXE60G - Nothing yet to report
Corsair Performance 3 Series CSSD-P3128GB2-BRKT - Both drives are throwing failures but drives are still functional
Crucial RealSSD C300 CTFDDAC064MAG-1G1 - Nothing yet to report
SAMSUNG 470 Series MZ-5PA128/US - Nothing yet to report
Intel 510 Series (Elm Crest) SSDSC2MH120A2K5 - One drive failed completely July 24 @ 9:14am and second drive is throwing failures but has yet to fail.
Intel X25-M SSDSA2MH160G2K5 - Nothing yet to report
Kingston SSDNow V+ Series SNVP325-S2B/128GB - a couple minor sector failures on one of the drives.


-------- End of line ----------



Any further questions?

[johnw]

Any further questions?

Yeah, what are you testing and how are you testing it?

[Ao1]

Any further questions?

Are the SF drives untrottled?

[sergiu]

Any further questions?

What do you understand by "throwing failures but has yet to fail"? Could you give us some details about what is happening? Also, how did Intel drive failed? did it reported SMART errors that would have indicated an imminent failure?

[B.A.T]

Todays update;

147.1145 TiB
494 hours
Avg speed 89.77 MiB/s.
AD gone from 17 to 15.
P/E 2564.
MD5 OK.

[Anvil]

184.88TB Host writes
MWI 1 (stopped moving)
Reallocated sectors : 6

MD5 OK, 35.62MiB/s avg (~17 hours)