SSD Write Endurance 25nm Vs 34nm

[B.A.T]

BAT

Change from HEX to DEC in CDI, (function -> Advanced -> raw values)

11h = 17

--

123.49TB Host writes
MWI 32

No other changes.

Fixed
Thank you!

[Ao1]

vapor for the V2 the write speeds varied depending on loop cycle. It got faster as the cycle extended. I would say that 45MB/s was a reasonable average.

John, I appreciate the Samsung is a bit of a nightmare to monitor.

The other thing that will be interesting is how/ if that drive manages to even wear out by rotating static data at the speeds it is writing.

[Vapor]

vapor for the V2 the write speeds varied depending on loop cycle. It got faster as the cycle extended. I would say that 45MB/s was a reasonable average.

Thanks! This is why I'm not going off the screenshots...speed can vary within a loop or across loops and I'd think each person has a better grasp of the typical/average speeds than what a single snapshot says

[B.A.T]

2241 GiB written, Wear Leveling Count and Percentage of the rated lifetime used has gone from 100 to 99.

[bulanula]

So eager to see how the C300 performs. It will surely beat the M4 but will it beat the Intel ones ?

[B.A.T]

Only time will tell us that. It'll be an exciting competition

[Ao1]

Thanks! This is why I'm not going off the screenshots...speed can vary within a loop or across loops and I'd think each person has a better grasp of the typical/average speeds than what a single snapshot says

Maybe data per day rather than MB/s would give a better, more even picture.

EDIT - When I say avg, that is based on what it would typically be when I looked at it. It was slower, but never much faster. (Not with compressible data anyway)

Anvil, a avg write speed counter would be quiet handy

[bulanula]

Indeed we also should be able to see the speeds. A drive lasting a decade at 10 MB / s is useless to most people.

[Anvil]

Will have a look at the avg speed.

My "X25-V" starts off at ~40MB/s and at the end of the loop its in the low 30s.
I do run the toolbox about every 24 hours, if I don't, it runs a little bit slower.

[Hopalong X]

Anvil and One Hertz

Per the Kingston and Intel.

Wondering if you have checked read/write speeds lately? If so maybe post them.
Just wondering if they have maintained or dropped from the heavy usage.

I haven't seen them posted lately.
Thanks
Mike

[One_Hertz]

My speeds are the same as when they were at the start... It has not slowed down.

[Hopalong X]

Thanks One Hertz.

[Vapor]

Based on write-days estimation, here's some stats:

X25-V: ~20 days from now until MWI = 0, based on 33.5MB/s average writes. Projected MWI Exhaustion in ~65 days (from a new state).
320 40GB: ~7-8 days from now until MWI = 0, based on 42MB/s average writes. Projected MWI Exhaustion in ~55 days (from a new state).
V2 40GB: projected MWI Exhaustion in ~55 days (from a new state) if LTT were disabled, based on 45MB/s average writes.
Samsung 470: projected MWI Exhaustion in ~6 days (from a new state), based on 140MB/s average writes.

Will post up charts when I get better/confirmed write speeds...all but the V2-40GB are just my guesses based on screenshots in this thread. Of course, it is also too early to tell if the speeds seen so far are sustainable until the end of a drive's life (and we don't even know if MWI = 0 means imminent death).

[johnw]

26.387 TiB , 79 hours, sa177: 56/56/2174

30.895 TiB , 91 hours, sa177: 48/48/2616

You can get average write speed by taking the difference between this and the previous TiB and hours.

If I need to stop the test for more than a few minutes, I will note it in my next update, so the speed calculation can be adjusted. Otherwise, you can just assume the test was running continuously for all the hours since the last update.

[B.A.T]

4,6361 TiB, 19 hours, Wear Leveling Count and Percentage of the rated lifetime used has gone from 99 to 98.

[bulanula]

Nice results. Eager to see after MWI = 0 point !

[Anvil]

125.2TB Host writes
MWI 31

Still at 6 reallocated sectors.

--

Some other stats
5.6229TB written over 416 loops 48.9 hours ~33.5MB/s on avg.
(includes 22,623.89 MB of random writes)

[B.A.T]

Well, that was the first 24h. At this point the M4 has written 6,0485 TiB. Wear Leveling Count and Percentage of the rated lifetime used has gone from 98 to 97.

[Ao1]

Will have a look at the avg speed.

My "X25-V" starts off at ~40MB/s and at the end of the loop its in the low 30s.
I do run the toolbox about every 24 hours, if I don't, it runs a little bit slower.

The V2 behaved the opposite. Started off slower and then cost faster as the loop progressed.

[B.A.T]

Same goes for my M4. It starts around 40MiB/s and tops off at around 92 MiB/s

[Anvil]

I'll have a look at it, it might have changed since I started doing manual trims.
The new version with the 5 second pause started running today as well.

Will be performing a manual MD5 test later today, so I'll monitor one loop for throughput at that time.

[Vapor]

30.895 TiB , 91 hours, sa177: 48/48/2616

You can get average write speed by taking the difference between this and the previous TiB and hours.

If I need to stop the test for more than a few minutes, I will note it in my next update, so the speed calculation can be adjusted. Otherwise, you can just assume the test was running continuously for all the hours since the last update.

Works out to 109.4MiB/s if 12hrs = exactly 12hrs. Also includes all the breaks/pauses in the program, not sure if they should be included in the average or not

Same goes for my M4. It starts around 40MiB/s and tops off at around 92 MiB/s

I had preliminary average of ~90MB/s for you...dropped it down to 75MB/s now but it's still preliminary.

In chart news, m4's data line was finally large enough for me to be able to click on it and add a trendline Considering how soon into testing it is, the trendline probably doesn't mean much right now, but it's there


Raw data graphs

Writes vs. Wear:


MWI Exhaustion:


Host Writes So Far:

(bars with a border = testing stopped/completed)


Normalized data graphs

Writes vs. Wear:
Attachment 117065

MWI Exhaustion:



Write-day data graphs

Write-days vs. Wear:

(preliminary until better average speeds are known)

[B.A.T]

6,0485 TiB in 24h = 73,41 MiB/s but I had 2 hours where the computer was down så 75 is probably spot on.

[One_Hertz]

165.4TB. 14%. 16 reallocated sectors.

41.8MB/s is my average write speed (calculated over the past 4 hours).

[johnw]

30.895 TiB , 91 hours, sa177: 48/48/2616

35.306 TiB, 103 hours, sa177: 41/41/2970