SSD Write Endurance 25nm Vs 34nm

[Cyber-Mav]

does toggle mode nand as in the plextor m3 have more durability than regular nand such as that in the crucial m4?

[Christopher]

does toggle mode nand as in the plextor m3 have more durability than regular nand such as that in the crucial m4?

Not necessarily; they're still governed by the same principles which limit endurance. With that said, the companies which make NAND and SSDs tend to find the best flash out of their fab facilites to put in their own drives. Intel/Micron(Crucial) and Samsung make flash, and this flash can go either in their own products, or get sold to other companies. Usually, they keep the best stuff for themselves. (Note that Toshiba does make flash, and does make drives, but they primarily are OEM units and some enterprise oriented offerings, but the same still probably applies to them)

The Plextor M3P is a fine product, but it might actually have less endurance than the M4. Perhaps we'll find out soon, I think I'm going to test one.

[Christopher]

Samsung 830 256GB Day 67

1,603,779.67GiB
1,566.19 TiB

288.07 MB/s Average
1590 Hours

6967 Wear Leveling Count
MWI 1

6/0 Erase/Program Fail
Used Reserved Block Count: 12/24572 sectors

I stopped the 830 and migrated to RC, all is well.

[canthearu]

Vertex 4 looks dead:



Smart values at end of life:



Drive has now disappeared from windows. I'll attempt to resurrect.

[Anvil]

[device is not ready] never was a good sign!

It's worth a bit of effort of course, it was showing the signs of a dying drive...

At 61% remaining life it is quite the disappointment. (if this is it)

[canthearu]

Well, I got it running again. It is now Zombie Vertex 4.

A couple of restarts got it re-detected.

Remaining life is probably closer to 31% given the drive life reset when I upgraded firmware. In any case, not good at all.

[canthearu]

Zombie Vertex 4 stopped again ... same position and error.

Wondering if I should try secure erase.

[Christopher]

Well, I got it running again. It is now Zombie Vertex 4.

A couple of restarts got it re-detected.

Remaining life is probably closer to 31% given the drive life reset when I upgraded firmware. In any case, not good at all.

This sounds very familiar... See the two VTs and the Octane...

[Anvil]

~30% is better, nevertheless, Remaining Life is there for a reason.

What about giving it a some extra over-provisioning?

[B.A.T]

Todays update:

m4
938.1534 TiB
3953 hours
Avg speed 74.35 MiB/s.
AD gone from 77 to 74.
P/E 16162.
MD5 OK.
Reallocated sectors : 00


Intel X25-M G1 80GB
647.67 TiB
23265 hours
Reallocated sectors : 161
Available Reserved space: 90
MWI=124
MD5 =OK
42.71 MiB/s on avg


Intel X25-E 64GB
113.14 TiB
352-30= 322 hours
Reallocated sectors : 0
Available Reserved space: 100
MWI=99
MD5 =OK
101.12 MiB/s on avg

[Ao1]

So much for the marketing gimmick otherwise known as the Ndurance 2.0. Is Ndurance just a throttling process? Surely someone has a grasp by now on what is going on with the V4 sustained write speeds?

[canthearu]

So much for the marketing gimmick otherwise known as the Ndurance 2.0. Is Ndurance just a throttling process? Surely someone has a grasp by now on what is going on with the V4 sustained write speeds?

That is what I've been thinking for the last week.

Ndurance 2.0 = almost lasts as long as the NAND is designed for.

What is even more confounding is how they are making the marvell controllers perform so poorly for endurance when the they have typically experienced perfectly fine endurance control.

[Christopher]

It's all about the flash. Period.

[Ao1]

For sure it would appear from testing in this thread that the grade of NAND used in Intel/ Micron/ Samsung SSD’s is superior to what others like $OCZ can buy on the open market, but that does not explain everything. So far the 520 has behaved much better than any other SF based drive. The premature block failures of the Octane/ Petrol drives are occurring before the P/E cycles have hardly been touched. It’s almost like something is accelerating/ inducing damage to the NAND. (Maybe it’s the overclock).

So what gives with the V4? Is it a form of throttling that causes write speed to drop by half? Someone must know, but no one is saying anything. If it was an Intel/ Micron/ Samsung/ Toshiba product the world and his dog would be screaming about it from the rooftops.

[Hopalong X]

So what gives with the V4? Is it a form of throttling that causes write speed to drop by half? Someone must know, but no one is saying anything. If it was an Intel/ Micron/ Samsung/ Toshiba product the world and his dog would be screaming about it from the rooftops.

They are still sceaming trying to get their Vertex 2's to work.

[Anvil]

Kingston SSDNow 40GB (X25-V)

982.99TB Host writes
Reallocated sectors : 05 45 up 2
Available Reserved Space : E8 99
POH 8806
MD5 OK

33.84MiB/s on avg (~48 hours)

--

Intel 330 120GB

20.37TB Host writes
Reallocated sectors : 05 0
Available Reserved Space : E8 100
[F1] Total LBAs Written 667628
[F2] Total LBAs Read 14536
[F9] Total NAND Writes 14696GB
POH 81
MD5 OK

122.49MiB/s on avg (~48 hours)

[Ao1]

Kingston SSDNow 40GB (X25-V)

982.99TB Host writes
Reallocated sectors : 05 45 up 2
Available Reserved Space : E8 99
POH 8806
MD5 OK

33.84MiB/s on avg (~48 hours)

6 more days until the PB!

[Anvil]

The 40GB drive has amazed me!
(got 3 more of them and I'll be keeping them, all of them are in use as boot drives with ~1TB of data written)

The 330 is rather nice as well, when prices settle it should be a great low cost alternative.

I need more time with the V4s, what I can say is that the 256GB drive does not behave like the 128GB.

[Anvil]

One more reallocated sector on the X25-V, it is now 46.

MWI on the 330 changed to 99, should be able to find out exactly when using the log-file.
(current values are 21.81TB Host writes, NAND writes are at 15729GB)

[Christopher]

Samsung 830 256GB Day 68

1,628,111.97GiB
1,589.95 TiB

279.63 MB/s Average
1615 Hours

7072 Wear Leveling Count
MWI 1

6/0 Erase/Program Fail
Used Reserved Block Count: 12/24572 sectors

It's getting... a little slow.

[canthearu]

Just tried a secure erase on the vertex 4, but no dice ... the drive continues to die after a loop or two, disconnecting with the same error experienced before.

At least it seems that the drive lets you get your data off it though ... I guess it is a small blessing.

Vertex 4 being removed from testing with disgust! The end.

Gonna have to start thinking of a new drive to test.

[Hopalong X]

This thread and Anvils app thread got "Sticky" designation.

About time.

[AMD_Freak]

Woot and I see that Anvil even got the MOD tag ...gratz one of the few threads I make a point at looking at daily.

[B.A.T]

Congratz Anvil

My update comes later tonight. This nice weather must be enjoyed first.

[B.A.T]

Todays update:

m4
945.7750 TiB
3983 hours
Avg speed 74.36 MiB/s.
AD gone from 74 to 69.
P/E 16292.
MD5 OK.
Reallocated sectors : 00


Intel X25-M G1 80GB
652.10 TiB
23295hours
Reallocated sectors : 161 to 162
Available Reserved space: 90
MWI=124
MD5 =OK
42.77 MiB/s on avg


Intel X25-E 64GB
123.52 TiB
382-30=352 hours
Reallocated sectors : 0
Available Reserved space: 100
MWI=99
MD5 =OK
101.08 MiB/s on avg