Printable View
256K chunks is not unlikely, it might be the actual block size for that NAND?
@adrianu
If the 830 is running with TRIM enabled you should be OK without reserving extra spare area. (over-provisioning)
It depends on what the drive is used for, normal usage you should be fine using the full capacity whether TRIM is enabled or not.
Without TRIM I'd personally set user-size to 110GB or thereabouts but it isn't required of course.
Well, I spoke too soon. I can't really figure out what the relationship is with the MOBIs smart value to writes. It could be some kind of bit error counter.
It is rather annoying that two f/w updates got issued within days of starting the test. Still I have separate log files for each f/w version, which I have combined together for the graphs below.Quote:
Originally Posted by Anvil
Prior to the 1st f/w update the #177 steps due to the fact that it did not update without a reboot. The subsequent big jump in the total LBA’s was due to a temporary switch over from the 4K endurance load to the normal endurance load.
*The 177 value is to the power of 10 to improve the scale visibility.
http://img440.imageshack.us/img440/2790/fwupdates.png
Now I add some trend lines. It seems that the wear has been linear regardless of workload or f/w version. I thought the enterprise workload would kill the drive in no time, but in practice it has not degraded the drive any faster than the normal endurance workload. The big difference of course being the write speeds.
http://img850.imageshack.us/img850/10/trendlines.png
@Ao1
That is interesting, I expect the large cache plays an important role in relation to that the workloads results in the same wear.
So, even if the OS can't do write combining the drive is capable of doing so, I find that impressive as the test is full span.
I don't think a higher QD/Thread count would have changed anything but it would of course have to be tested.
@Christopher
There is a bit of interesting info on the Mobi out there Link
7.2.4 Total erase count
Total Erase Count means the summation of erase counts of all blocks in SSD.
In general, the maximum value of 32GB SSD's Total Erase Count is 3200M and if the Total Erase Count exceeds the
value, it says that the SSD is completely wore - out.
Maximum Total Erase Count differs from vendors of NAND FLASH Chip and the capacity of SSD.
7.2.5 Total erase count current value
This value becomes 100 when Total Erase Count is equal to 0 and becomes equal to 0 when Total Erase Count is
maximum value.
If the value is equal to 0, it means that the SSD is completely wore-out.
7.2.6 Total erase count worst value
This value becomes 100 when Total Erase Count is equal to 0 and becomes 0 when total erase count is 90% of maximum
value.
Users should be warned when this value becomes equal to 0.
7.2.7 Total erase count raw value
This value means the summation of erase counts of all blocks in SSD.
You'll find it on page 17 and on-wards.
So, the current value acts as MWI
Over the course of the test, the 830 has used 1 PE cycle per 23GB. But if you look at just the last two days, that number is 58GB/PE cycle.
If the drive died right now, it would be a decent result in terms of PE cycles. But if the numbers are accurate, and it makes it pretty far, it will be a good story.
The Vertex Turbo is using 1 PE cycle for every 56GBs.
Samsung 830 64GB Update, Day 50
FW:CXM02B1Q
GiB Written:
329588.84
TiB written:
321.8641
Avg MB/s
132.36
PE Cycles
15494 up from 15300 <--- this is not a lot, it used to do 370+ a day.
Reallocated Sectors
45056
22 Blocks, up from 20!
1187 hours
https://www.box.com/shared/static/uy...5evmtbpje7.png
https://www.box.com/shared/static/z4...jvbdf9u7jq.png
OCZ Vertex Turbo 64 Update Day 6
TiB 35062.37
GiB 34.2406
Avg MB/s 83.09
Avg Erase Count
621 up from 496
MWI 88 down from 91
1 Program Fail
19 Erase Fail (up from 16)
0 Read Fail
129 Hours
https://www.box.com/shared/static/cl...n51iof4xp4.png
https://www.box.com/shared/static/f2...5stnjm9bm6.png
Host writes: 15,236 GiB
MWI: 11
P/E Cycles: 3,232
POH: 401
Relocated sectors: 8,192
MB/s: 5.95
Here is a graph showing writes per day and changes in the P/E count. Once the P/E cycles updated correctly the drop in the P/E count is linear to the write activity.
http://img593.imageshack.us/img593/5...ingpecount.png
I need to run some benchmarks on the 830 -- While average writes are fast, and getting faster, I still don't know if reads are back to where they should be.
It's just that, damn... the 830 is fast right now. Really fast. It's creeping up on 134MB/s avg (and instantaneous readings throughout the loop are in the 143 - 147MBs range). That's faster than the Mushkin Chronos D at 46 percent compressibility.
I'm considering changing from the X25-V as the bootdrive to the 830 64GB on 2 or 3 computers, I'm still waiting for the Intel 520 though :)
(those are all 3Gb/s setups but still I find it worth upgrading)
I’m waiting for the 830 to get to MWI 1 before changing over to a 830 for my OS drive. I'm looking forward to it although the 60GB drive I have is a bit too small. I might get an Intel 520, but only if Intel has managed to improve performance with compressed data. If not I’m not interested.
Excuse me if some of these questions are a bit dense, just trying to grasp SSD endurance testing...Quote:
Originally Posted by Ao1
1) if you've only written 15TiB but the MWI is already at 11 isn't that very bad? (Assuming MWI=100 initially and 0 means that one should replace the drive [according to the manufacturer]). So assuming that MWI is decreasing linearly w.r.t host writes you'll only have a "official" life of only ~17TiB?
2) why is your MB/s so low compared with Christopher's for the same drive? I'm probably missing something here..
Ao1 is doing the test differently than I am. He is doing full span 4K writes which dramatically increase write amplification and as a result is much slower. The whole drive is full, and writing each additional 4K chunk requires a lot of effort on the part of the drive.Quote:
Originally Posted by lm358
My 830 is being tested like all the other drives are. It has 18.5GB of static data, and writes on the remaining space until the minimum amount of free space is reached (normally around 12GB). So around 30GB gets written each loop, at which point the files get deleted and the drive starts again (after writing 100MB of random writes). Repeat ad nauseum.
Kingston SSDNow 40GB (X25-V)
658.98TB Host writes
Reallocated sectors : 05 21
Available Reserved Space : E8 99
POH 5944
MD5 OK
--.--MiB/s on avg (~-- hours)
--
Corsair Force 3 120GB
01 120/50 (Raw read error rate)
05 2 (Retired Block count)
B1 42 (Wear range delta)
B6 1 (Erase Fail Count)
E6 100 (Life curve status)
E7 10 (SSD Life left)
E9 771069 (Raw writes) ->753TiB
F1 1026074 (Host writes) ->1002TiB (~22TiB left until 1PiB is reached)
MD5 OK
---.--MiB/s on avg (~-- hours)
power on hours : 2917
--
System was frozen again tonight, I discovered the issue within an hour so there's not much missing.
Spent some time figuring out if there are changes to the system but I can't think of anything...
I have applied the LPM fix, lets see if it behaves.
Todays update.
Kingston V+100
346.7130 TiB
2034 hours
Avg speed 26.26 MiB/s
AD still 1.
168= 1 (SATA PHY Error Count)
P/E?
MD5 OK.
Reallocated sectors : 00
http://www.diskusjon.no/index.php?ap...tach_id=476886
Intel X25-M G1 80GB
277,65 TiB
20488 hours
Reallocated sectors : 00
MWI=126 to 124
MD5 =OK
49.22 MiB/s on avg
http://www.diskusjon.no/index.php?ap...tach_id=476888
m4
301.3858 TiB
1112 hours
Avg speed 81.57 MiB/s.
AD gone from 187 to 180.
P/E 5307.
MD5 OK.
Reallocated sectors : 00
http://www.diskusjon.no/index.php?ap...tach_id=476887
SSDLife Pro is still not working so you just got ASU this time.
I noticed you've restarted the Kingston V+100, is it failing on a daily basis?
Samsung 830 64GB Update, Day 51
FW:CXM02B1Q
TiB Written:
340160.01
GiB written:
332.1875
Avg MB/s
132.97 by 7:30PM, the drive has written 9000GB -- which is awesome
PE Cycles
15667
Reallocated Sectors
45056
22 Blocks, holding steady
1210 hours
https://www.box.com/shared/static/ig...6g4k6e99hz.png
https://www.box.com/shared/static/2q...yrz4f4gar6.png
OCZ Vertex Turbo 64 Update Day 7
TiB 40.1624
GiB 41126.35
Avg MB/s 82.02
Avg Erase Count
736
MWI 86 down from 93
1 Program Fail
20 Erase Fail (up from 19)
0 Read Fail
152 Hours
https://www.box.com/shared/static/td...qpyp4r1txi.png
https://www.box.com/shared/static/x9...1lz24a2qve.png
That Kingston... I ended up with what I'm certain is almost identical drive, and it's not spectacular either. I tried to get an Imation S Class slc 32gb drive a while back, and I recieved an M-Class J.Micron controlled Toshiba NAND 64GB drive. It's not stellar, and has the same smart info as the Imation MTRONs. The Toshiba controller in the Kingston is just a JMicron, right?
Incidentally, the next time I take the endurance rig down for maintenance, I'm tossing in one of the MTRON 16GBs (whichever drive I can shoehorn in there). I'm probably just do biweekly updates with it, but I'm curious what will happen. I was thinking 4GB of static data (perhaps the Windows7 Install DVD). If it can hold 50MB/s plus writes, I think something interesting should happen in 2013.
Check out these 830 Benches.
The write speed is back to normal... but the reads are awful.
https://www.box.com/shared/static/xo...nvpjqmsezd.jpg
SATA III is working properly, too. I double and triple checked.
I took before and after secure erasure benches, but they both look like this.
Hey, guys.
I'm sorry if I'm diverting the cause of this thread, or even thread crapping, but I feel like I need to ask something.
Seeing the results of your laborious review of SSD failure rate,
I am wondering whether you think that a 128GB C300 would be a better choice in terms of longevity than a 128GB M4.
I am in the market for such drive and plan on buying it later today.
Umm, I don't think the answer is obvious. The C300 is a fine drive, but so is the M4. I don't think there is a reason to avoid the M4... the first M4 that B.A.T. tested made it well past 700,000GB of writes.Quote:
Originally Posted by Track
That one was a 64GB model, so the 128 could maybe do double that? Either way, drives do die on occasion -- but practically never from wearing out the flash. That's something that can happen here in testing, but it's really not a possibility for a normal consumer workload over five years (or longer).
For example, my main desktop writes about 10GB a day. Yesterday my Samsung 830 I'm endurance testing wrote 1100x that amount. It's been doing that for over 50 days now. The Intel X35-V in the test has been writing as fast as it can since last May. Your drive spend only a couple of minutes a day writing data to the drive.
You may also wish to consider the Plextor M2P and Corsair Performance Pro 128 GB drives. They use the same Marvel controller as the M4 but with 32nm Toshiba Toggle NAND.
Either way, the M4 and C300 are both fantastic and generally well reguarded in terms of reliability. The C300 looks a little slow in comparison now to it's contemporaries, but is very fast none the less.
LASTLY:
We're trying to kill the SSDs, in case you weren't really aware. We're not "testing failure rate" ; on a long enough time line, the failure rate for all the drives in the test should be 100%. The test started as a way to determine if 2xnm NAND was as "good" as 3xnm NAND, but it's a good way to test and understand other aspects of SSDs.
Essentially, we kill drives (on purpose), and we're pretty good at it.
Host writes: 15,752 GiB
MWI: 7
P/E Cycles: 3,376
POH: 401
Relocated sectors: 8,192
MB/s: 6.25
I just checked my 830. I stopped the endurance test but left the test file intact, so this is on a full drive that has had no chance to recover from 4K random writes. AS SSD crashed half way through, which I guess was due to the fact that the drive had no space to write.
Next I deleted the test file and ran the test again. Impressive. :cool:
http://img217.imageshack.us/img217/4...dpredelete.png
http://img208.imageshack.us/img208/8...postdelete.png
Ao1,
Your drive looks impressive. Mine... not so much.
Yeah, I secure erased mine -- and it didn't help. The writes are very impressive, but my reads are in the toilet and I don't really have an explanation.
Now, the reads used to be better while the writes were bad, but now the opposite is true. It's currently over 134MBs avg, and one out of every three loops is 160MBs instantaneous while the previous two are in the high 140's.
It fails betwheen 24-36-72 hours and increasingly the last 3 weeks. :shakes:Quote:
Originally Posted by Anvil
The Kingston V+100 got a Toshiba controller labeled T6UG1XBG.Quote:
Originally Posted by Christopher
They are booth good SSD. Go for the cheapest and fastest one. You can't go wrong with them. :up:Quote:
Originally Posted by Track