SSD Write Endurance 25nm Vs 34nm

[DragoonXX]

Forget the computer, YOU will be old and slow before you ever write too much to the M4...

I agree on this too. just checked the writes on my 2 x-25V drives in a raid 0- one has written 0.04TB, and the other has written 0.03TB in the last 24 and a half days, they are now sitting at 1.36TB with 10285 power on hours, and 1.23TB at 10252 power on hours- as of last update- the x-25v comparable had written 375.29TB...if my usage pattern doesnt change- I'd be looking at...about 320 years before it died *assuming it would die at 375.29TB*

[Christopher]

I agree on this too. just checked the writes on my 2 x-25V drives in a raid 0- one has written 0.04TB, and the other has written 0.03TB in the last 24 and a half days, they are now sitting at 1.36TB with 10285 power on hours, and 1.23TB at 10252 power on hours- as of last update- the x-25v comparable had written 375.29TB...if my usage pattern doesnt change- I'd be looking at...about 320 years before it died *assuming it would die at 375.29TB*

Exactly. In fact, I was writing much more a day, more like 10GB, but that was mostly just from leaving Chrome open with several tabs all day in addition to the Windows Live Mesh client I was using to sync important stuff across every SSD with my main system image (like 7 drives... very handy, all of them are up to date). Without those two programs, I'm writing about 1GB every 35 - 50 hours or so.

I'm beginning to think that modern controllers with super low WA and intelligent wear leveling means your drives will be obsolete before you ever get 25% of the way to running out of PE cycles. Who knows how long SATA will be around? 5 more years? 7? I'm sure it'll stick around awhile for HDDs, but I think that SSDs will begin to utilize another standard over the next couple years for better sequential speeds. But I still think that today's SSDs are going to be useful until motherboards don't have SATA anymore and that they'll still stack up well in certain performance and longevity parameters for years to come. Two X25-Vs are still going to be impressive two years from now. So I see the SSDs I have as long term investments, not something I'll get rid of when I upgrade my system. Much like a good sound card or excellent power supply, those are the parts you take with you.

[B.A.T]

Forget the computer, YOU will be old and slow before you ever write too much to the M4...

600TB over 10 years is 164GB a day. SSDlife reports that the system drive on my system has written .2GB in thirteen hours... it's an X25-E 32GB SLC, so it could last until the next ice age at that glacial pace of writes...

You are probably right there. My Intel X25-M 160GB will follow me to the grave. 6.8 TB written in 2 years.....

[Christopher]

Just toss 'em in the ground with me...

[deathman20]

You are probably right there. My Intel X25-M 160GB will follow me to the grave. 6.8 TB written in 2 years.....

Yeah my 2x 80Gig G2's I had was 1TB over a 2 year period. Guess everyone is so worried that your going to kill the drive so quickly with writes they worry. Obviously its not the case, and I even have stressed this. This obviously backs up the data, and then some. The drive can obviously take on writes... alot of them. Heck if you get over 600TB with that 64Gig my 128Gig might be good for twice that if it died at that. Yup a loooooooooong time before I worry about the drive going oops dead. I know I'll upgrade it well before then, the search for more speed

[bulanula]

Any update on the Sammy ? Did johnw send it off already to Ottawa ?

[Christopher]

Any update on the Sammy ? Did johnw send it off already to Ottawa ?

Yes, we need updates on the F40-A and the 470's current disposition...

[One_Hertz]

Any update on the Sammy ? Did johnw send it off already to Ottawa ?

It is on its way to me via USPS.

[Christopher]

The Samsung could be a treasure trove of info... I'm very curious as to what's actually going on inside.

[Anvil]

Kingston SSDNow 40GB (X25-V)

376.80TB Host writes
Reallocated sectors : 10
MD5 OK

33.46MiB/s on avg (~71 hours)

--

Corsair Force 3 120GB

01 90/50 (Raw read error rate)
05 2 (Retired Block count)
B1 59 (Wear range delta)
E6 100 (Life curve status)
E7 68 (SSD Life left)
E9 128091 (Raw writes)
F1 170626 (Host writes)

109.93MiB/s on avg (~71 hours)

power on hours : 489

Just to make it clear

I have not upgraded the OROM on this MB (ASRock) it is still running the original bios (1.60) which means that it's OROM 10.6.0.1091 and I am running the latest official RST Kit which is 10.6.0.1002/1022

The one and only thing I have changed is that I have disabled Power savings.
No C states, nothing.

[Christopher]

Okay Anvil, I'll try disabling C states. So your ASRock doesn't have speed step/EIST/Turbo enabled either? So it's just idling at 3.3GHZ all the time (Or 3.4 for a 2600K)?

[Anvil]

That is correct, it's idling at 34x. (no Turbo)
(All power saving related options are disabled and that includes SpeedStep etc)

I'll post the screenshots when I can get to the bios.

Power Options (Control Panel) is set to High Performance as well.

[Christopher]

Maybe I'll try that, but just set the multiplier much lower.

[Christopher]

Mushkin Chronos Deluxe 60 Update, Day 16

05 2
Retired Block Count

B1 9 (Down from 10)
Wear Range Delta

F1 149566
Host Writes

E9 115316
NAND Writes

E6 100
Life Curve

E7 40
Life Left

Average 126.62MB/s Avg
RST drivers, Intel DP67BG P67

372 Hours Work (23hrs since the last update)
Time 15days 12hours

12GiB Minimum Free Space 11160 files per loop

SSDlife expects 9 days to 0 MWI


The Wear Range Delta continues to decrease. I'm not entirely sure whether wear range delta is a percentage figure (9% difference between most used and least used) or if it's a number of PE cycles (that would be impressive if the difference between all cells is only 9PE cycles).

[Computurd]

Forget the computer, YOU will be old and slow before you ever write too much to the M4..

LOL- well said!

[devsk]

AD gone from 02 to 254. The counter has been started over again from 255. Wonder how far this will go.....

Thats just a uint8_t counter rotating! Has no meaning anymore!

[DragoonXX]

I have this feeling that I'll end up with like 4 SSDs in my next comp (2 new ones +my current X-25v for scratch disks)
I've been watching this thread and getting more and more impressed with endurance as time has been going on. last semester one of my instructors had asked me if I could put together some ideas for replacement systems for the department (drafting)- at the time, I was leery about suggesting SSDs due to the 4 year replacement cycle (and then later finding out that SSD+ deep freeze=bad) and the instructor and me both ended up joking around saying that after 4 years the SSDs would have holes burned through them from such heavy wear and tear.

Enter May and this thread starts, start watching it, start getting impressed, I kept that instructor in touch with the progress- talked to him before the semester started and he ordered in an OWC drive for his macbook pro

[B.A.T]

Thats just a uint8_t counter rotating! Has no meaning anymore!

I know. It's the 3rd time the counter has been reset. According to SSDLife I'm on 358% and increasing

[etlamo]

Since SMART formula of 0xCA is
100-100*X, X ~= WearLevel / MaxWearLevel (Max = 3000)
and byte rotating:

100*(1-X) == 0 on Wear level = 3000
100*(1-X) == -256 on Wear level = 10680
100*(1-X) == -512 on Wear level = 18360 ( Spoiler )
100*(1-X) == -768 on Wear level = 26040
100*(1-X) == -1024 on Wear level = 33720
...
So far it could reset the counter only 2 times, at 3000 and 10680.

[sergiu]

The Wear Range Delta continues to decrease. I'm not entirely sure whether wear range delta is a percentage figure (9% difference between most used and least used) or if it's a number of PE cycles (that would be impressive if the difference between all cells is only 9PE cycles).

Most likely it is measured in % of total cycles. Probably the algorithm does not make any big effort to level the wearing until a certain wear has been reached, like 50%

[B.A.T]

Since SMART formula of 0xCA is
100-100*X, X ~= WearLevel / MaxWearLevel (Max = 3000)
and byte rotating:

100*(1-X) == 0 on Wear level = 3000
100*(1-X) == -256 on Wear level = 10680
100*(1-X) == -512 on Wear level = 18360 ( Spoiler )
100*(1-X) == -768 on Wear level = 26040
100*(1-X) == -1024 on Wear level = 33720
...
So far it could reset the counter only 2 times, at 3000 and 10680.

Let me correct my self. It's the 3rd time it's started over again

[Anvil]

Kingston SSDNow 40GB (X25-V)

378.06TB Host writes
Reallocated sectors : 10
MD5 OK

33.35MiB/s on avg (~83 hours)

--

Corsair Force 3 120GB

01 90/50 (Raw read error rate)
05 2 (Retired Block count)
B1 58 (Wear range delta)
E6 100 (Life curve status)
E7 67 (SSD Life left)
E9 131350 (Raw writes)
F1 174961 (Host writes)

109.84MiB/s on avg (~83 hours)

power on hours : 500

Notice that avg sped has dropped and Wear Range Delta has decreased as well.
There is clearly a connection between those factors. (meaning static wear leveling makes the drive run somewhat slower)

[B.A.T]

Todays update
m4
625.3004 TiB
2279 hours
Avg speed 91.07 MiB/s.
AD gone from 254 to 250.
P/E 10889.
MD5 OK.
Still no reallocated sectors


Kingston V+100
135.2691 TiB
530 hours
Avg speed 75.95 MiB/s.
AD gone from 68 to 61.
P/E ?.
MD5 OK.

[Computurd]

welcome etlamo

[Ao1]

Kingston SSDNow 40GB (X25-V)

Notice that avg sped has dropped and Wear Range Delta has decreased as well.
There is clearly a connection between those factors. (meaning static wear leveling makes the drive run somewhat slower)

Anvil I’m intrigued by this. AFAIK the SF controller is supposed to prevent the delta between the least-worn and most worn block from accumulating more than a few % of the maximum lifetime wear rating of the Flash memory.

I believe it is calculated something like this:

 Wear Range Delta = [(MW - LW) / MRW] x 100
 MW = P-E Cycles experienced by Most Worn block
 LW = P-E Cycles experienced by Least Worn block
 MRW = Max Rated Wear = P-E Cycle rating for the Flash memory

Your Delta is 58. If I assume a P/E rating of 5,000 the difference between least and most worn is significant and certainly well in excess of what the controller is designed to prevent.

 58 = [(2,900) / 5,000] x 100

Or 3,000 P/E

 58 = [(1,740) / 3,000] x 100

I suspect that static wear levelling can only occur when the drive is in idle mode. Most likely the controller cannot accept any other commands whilst it moves blocks around and flushes out any invalid data contained within the static data block.

Between post 2,060 & post 2,072 did your F3 have any power on idle time?

If static data cannot be rotated whilst the drive is active it means that the endurance app is inducing wear well above the rate that would occur if the data could be rotated.

Btw I believe that SF drives will issue a SMART trip once the reserve block count drops below the minimum allowable threshold.