Page 25 of 220 FirstFirst ... 15222324252627283575125 ... LastLast
Results 601 to 625 of 5494

Thread: SSD Write Endurance 25nm Vs 34nm

  1. #601
    Moderator Anvil's Avatar
    Join Date
    Apr 2008
    Location
    Norway
    Posts
    2,838
    Quote Originally Posted by B.A.T View Post
    Hello everyone!
    ...
    B.A.T


    Great having a "contest" on the Crucial's
    -
    Hardware:

  2. #602
    Admin Vapor's Avatar
    Join Date
    Feb 2005
    Location
    Ann Arbor, MI
    Posts
    13,107
    What are people expecting the outcome between the C300 and the M4?

    Should I open my C300 and ensure it's 34nm? I know there's been no controversy, but 34nm supply has to dry up some time....

  3. #603
    Xtreme Addict
    Join Date
    Nov 2005
    Location
    Where the Cheese Heads Reside
    Posts
    2,173
    Unless they continue making 34nm chips and just charge more for it? I mean the C300's are going for more than the M4's of similar size from what I've been seeing.

    Ok just ran on my drive wow shocked at the results
    Intel SSD1 Smart.pngIntel SSD2 Smart.png

    21 months of use.... I would of figured I would of burned through it more!
    Last edited by deathman20; 06-29-2011 at 05:41 PM.
    -=The Gamer=-
    MSI Z68A-GD65 (G3) | i5 2500k @ 4.5Ghz | 1.3875V | 28C Idle / 65C Load (LinX)
    8Gig G.Skill Ripjaw PC3-12800 9-9-9-24 @ 1600Mhz w/ 1.5V | TR Ultra eXtreme 120 w/ 2 Fans
    Sapphire 7950 VaporX 1150/1500 w/ 1.2V/1.5V | 32C Idle / 64C Load | 2x 128Gig Crucial M4 SSD's
    BitFenix Shinobi Window Case | SilverStone DA750 | Dell 2405FPW 24" Screen
    -=The Server=-
    Synology DS1511+ | Dual Core 1.8Ghz CPU | 30C Idle / 38C Load
    3 Gig PC2-6400 | 3x Samsung F4 2TB Raid5 | 2x Samsung F4 2TB
    Heat

  4. #604
    Registered User Hopalong X's Avatar
    Join Date
    Jun 2011
    Posts
    87

    Thumbs up

    Just saying hello for now.
    I've been lurking and reading the thread the last few weeks.
    I finally got the registration button to work on the forum.

    I bought my first SSD about the same time the testing started.
    Went with the Intel 320 120GB.
    The longer the testing goes the more I like it.
    Enjoying the thread immensly and it looks like more fun to come.

  5. #605
    SLC One_Hertz's Avatar
    Join Date
    Oct 2004
    Location
    Ottawa, Canada
    Posts
    2,953
    157.6tb. 18%.

  6. #606
    Registered User Khoral's Avatar
    Join Date
    Sep 2009
    Location
    Noisy le Sec, France
    Posts
    78
    Only posting to say thanks to all the testers

    This thread is a real mine of information
    Sorry for my bad english, I'm trying to improve it


  7. #607
    Xtreme Enthusiast DragoonXX's Avatar
    Join Date
    Mar 2008
    Location
    Alberta Canada
    Posts
    631
    Quote Originally Posted by deathman20 View Post
    Unless they continue making 34nm chips and just charge more for it? I mean the C300's are going for more than the M4's of similar size from what I've been seeing.

    Ok just ran on my drive wow shocked at the results


    21 months of use.... I would of figured I would of burned through it more!
    wow.. I've had my X-25V drives for closing in on 15 months, mine are reporting 8142 and 8138 hours power on time, and 0.99 and 1.09TB writes...which is probably close to about 1/10 what I thought I have done on them so far and they have a MWI of 98, with available reserved space of 100 and a 10 threshold
    Current System:
    eVGA 680i SLi "A2" P30 BIOS
    intel Core 2 Quad Q6600 (currently at stock)
    OCZ ReaperX 4GB DDR2 1000 (running at DDR2 800 Speeds with cas4)
    320GB Seagate 7200.10
    XFX 8800GT XXX 512MB (stock clocks)
    auzentech X-Fi Prelude
    PC Power and Cooling Silencer 750 Quad Copper
    Win XP Pro

  8. #608
    Xtreme Enthusiast
    Join Date
    Jun 2011
    Location
    Norway
    Posts
    609
    Thanks for the warm welcome

    Before we start the crucial test we need to agree on the config.
    - How much should the ssd be filled with static data? 40 vs 64 GB
    - What parameters should we use on Anvils app
    - How much random vs seq?

    Something i missed?
    1: AMD FX-8150-Sabertooth 990FX-8GB Corsair XMS3-C300 256GB-Gainward GTX 570-HX-750
    2: Phenom II X6 1100T-Asus M4A89TD Pro/usb3-8GB Corsair Dominator-Gainward GTX 460SE/-X25-V 40GB-(Crucial m4 64GB /Intel X25-M G1 80GB/X25-E 64GB/Mtron 7025/Vertex 1 donated to endurance testing)
    3: Asus U31JG - X25-M G2 160GB

  9. #609
    Moderator Anvil's Avatar
    Join Date
    Apr 2008
    Location
    Norway
    Posts
    2,838
    Quote Originally Posted by Vapor View Post
    What are people expecting the outcome between the C300 and the M4?
    Should I open my C300 and ensure it's 34nm? I know there's been no controversy, but 34nm supply has to dry up some time....
    I'm expecting them to exceed the 72TBW guarantee that Crucial has put on them, hopefully they'll get up there with the Intels.
    As there is more NAND they should match the Intels but as we know, the controller can make the difference

    I'm pretty sure there is 34nm NAND in it but maybe we should all open the drives that enter the Endurance test?

    Quote Originally Posted by Khoral View Post
    Only posting to say thanks to all the testers
    This thread is a real mine of information
    Appreciated

    Quote Originally Posted by deathman20 View Post
    Ok just ran on my drive wow shocked at the results
    21 months of use.... I would of figured I would of burned through it more!
    I'm not shocked at all, I've got plenty of SSD's that are low on writes.
    As long as it's used for normal tasks it just doesn't write as much as one thinks.

    My default setup is w/o the pagefile, system restore and the hibernation is off as well, these things will make a difference.

    But, as we all can see, you shouldn't really worry

    --

    119,49TB Host writes
    MWI 34

    Still at 6 reallocated sectors.
    -
    Hardware:

  10. #610
    Admin Vapor's Avatar
    Join Date
    Feb 2005
    Location
    Ann Arbor, MI
    Posts
    13,107
    Quote Originally Posted by B.A.T View Post
    Thanks for the warm welcome

    Before we start the crucial test we need to agree on the config.
    - How much should the ssd be filled with static data? 40 vs 64 GB
    - What parameters should we use on Anvils app
    - How much random vs seq?

    Something i missed?
    I'd like to know what johnw is using for his Samsung 470 64GB, that way all 64GB drives can be on the same test parameters.

    In the absence of that info, I'd say 20-24GB of static data (for a 64GB drive) and default values in Anvil's app.

  11. #611
    Xtreme Member
    Join Date
    Apr 2011
    Posts
    211
    Quote Originally Posted by Vapor View Post
    What are people expecting the outcome between the C300 and the M4?

    Should I open my C300 and ensure it's 34nm? I know there's been no controversy, but 34nm supply has to dry up some time....
    I fully expect the C300 to win. The M4 is just a cost cutting 25nm version of the C300.

    Please open the drive up and see what type of NAND it has. Usually 8GB chips ( or F64 ) are 34nm MLC and 16GB chips ( F128 ) are 25nm chips.

  12. #612
    Banned
    Join Date
    Jan 2010
    Location
    Las Vegas
    Posts
    936
    Quote Originally Posted by Vapor View Post
    I'd like to know what johnw is using for his Samsung 470 64GB, that way all 64GB drives can be on the same test parameters.

    In the absence of that info, I'd say 20-24GB of static data (for a 64GB drive) and default values in Anvil's app.
    I mentioned in my post that I put a ~40GB static file on the SSD. To be precise, it is 41992617078 Bytes (I imagine that is a typical amount of static data for a 64GB SSD) And Anvil's app and data is on the SSD. All settings in Anvil's app are the default, except that I checked the box about keeping running totals for GB written (option just added yesterday).

    For reference, the md5sum of the 42GB file is: 0d1c4ec44d9f4ece86e907ab479da280
    Last edited by johnw; 06-30-2011 at 05:47 AM.

  13. #613
    Xtreme Mentor Ao1's Avatar
    Join Date
    Feb 2009
    Posts
    2,597
    Now we have more drives joining in who will track all the data?

    This chart can go into the negative for wear out, but can only assume a negative MWI value based on average writes to date per MWI value.

    It's also hard to see all the data when it is small. I had to take out all the hard data as it was too small. The Y axis only represents TB for One_Hertz. With more drives it will get a bit harder, but it would be good if all drives could be on one chart.


    Untitled.png

  14. #614
    Admin Vapor's Avatar
    Join Date
    Feb 2005
    Location
    Ann Arbor, MI
    Posts
    13,107
    Quote Originally Posted by johnw View Post
    I mentioned in my post that I put a ~40GB static file on the SSD. To be precise, it is 41992617078 Bytes (I imagine that is a typical amount of static data for a 64GB SSD) And Anvil's app and data is on the SSD. All settings in Anvil's app are the default, except that I checked the box about keeping running totals for GB written (option just added yesterday).

    For reference, the md5sum of the 42GB file is: 0d1c4ec44d9f4ece86e907ab479da280
    Alright, got a file ready for myself, weighing in at 42,022,123,868 bytes. C300 64GB should start tomorrow so long as UPS sticks to their delivery date.

    I'm also going to test a SF-1200 drive now that the prospect of no-LTT has emerged (and if anyone wants to test a 25nm vs. my 34nm, let me know...easier to arrange testing and setup in pairs!). With a Sandforce back on the scene, I wanted to examine the compression settings in Anvil's app and see if any were suited to mimic 'real' data. With the discovery of the 233 SMART value, we can now see NAND writes in addition to Host writes, so if we can also write 'real' data we can kill two birds with one stone: see how long a drive lasts with 'real' use and how much the NAND can survive.

    So what did I do?

    First, I took two of my drives, C: and D:, which are comprised of OS and applications (C:) and documents (D:, .jpg, .png, .dng, .xlsx probably make up 95% of the data on it) and froze them into separate single-file, zero compression .rar documents. I then took those two .rar files (renamed to .r files...WinRAR wasn't too happy RARing a single .rar file) and ran them through 6 different compression algorithms: WinRAR Fastest RAR setting, WinRAR Normal RAR setting, WinRAR Best RAR setting, 7-zip Fastest LZMA setting, 7-zip Normal LZMA setting, and 7-zip Ultra LZMA setting. I then normalized the output file sizes.

    Doing this created two 'compression curves' showing how my real data responds to various levels of compression. My thinking being that if any of Anvil's data compressibility settings had similarly shaped and similarly sized (after normalization) outputs, it would be a good candidate to use to mimic real data and allow the use of 'real' data with SF testing. Real data != 'real' data; 'real' data is just the best attempt to generate gobs of data that walk, talk, and act like real data. A great candidate would be a generated data set that had a compression curve between the two curves from real data, across the entire curve.

    AnvilCompressionCurvesStart.png

    Once I had those curves mapped out, I made ~8GB files of each of the various settings with Anvil's app (0-fill, 8%, 25%, 46%, 67%, and 101%) and made curves for each of them.

    All put together, they look like this:
    AnvilCompressionCurves.png

    The green zone is where the potential candidates should show up. Only one candidate was in that range, however: 67%. Unfortunately, it fell out pretty aggressively with stronger compression algorithms. So I turned off the "Allow Deduplication" setting and generated another 8GB file and compression curve and it was a little better.

    While dedicated hardware can be magnitudes more efficient than a CPU with an intensive task, I do doubt the SF-1200 controller's ability to out-compress and out-dedup even low resource LZMA/RAR (R-Fastest and 7-Fastest), so the left-most part of the green zone is a stronger green as I feel that's the most important section of the curve. Unfortunately, I don't have the ability to get more granular compression curves at the low-end (left side) of the curve, so I'll have to make do with overall compression curves with just an emphasis on the low-end.
    Of all the data I have available it does look like 67% compression setting with "allow deduplication" unchecked seems to be the best fit for use as a 'real' data setting for when I start testing SF-1200. Hopefully anybody else who plans to test a controller with compression and deduplication will find this useful as well

  15. #615
    Admin Vapor's Avatar
    Join Date
    Feb 2005
    Location
    Ann Arbor, MI
    Posts
    13,107
    Quote Originally Posted by Ao1 View Post
    Now we have more drives joining in who will track all the data?

    This chart can go into the negative for wear out, but can only assume a negative MWI value based on average writes to date per MWI value.

    It's also hard to see all the data when it is small. I had to take out all the hard data as it was too small. The Y axis only represents TB for One_Hertz. With more drives it will get a bit harder, but it would be good if all drives could be on one chart.
    I volunteer to chart data...I'll start working through this thread for the data

  16. #616
    Xtreme Mentor Ao1's Avatar
    Join Date
    Feb 2009
    Posts
    2,597
    I'll PM you the excel file. If nothing else it has the raw data. It will be nice to join the peanut galley. I've been spending way too much time on that V2.

    If you run at 67% it will be interesting to see if you get throttled at some stage. I suspect you will.

    EDIT:

    How will you record writes? 231 & 241?
    Last edited by Ao1; 06-30-2011 at 09:49 AM.

  17. #617
    Moderator Anvil's Avatar
    Join Date
    Apr 2008
    Location
    Norway
    Posts
    2,838
    @Vapor
    Superb job on collecting compression data and yes, it's based on 7Zip Fast compression ratio. (could be Fastest, will check)

    Looking at my tests on the SF2 controller it couldn't keep up with the ratio that 7Zip Fast(est) produces, not sure how the SF1 handles vs the SF2.

    For reference, I 7Zipped one of my VM's earlier today (Windows Server 2008 R2, SQL-Server, + some data) and it ended up being ~50% of the original size using 7Zip Fastest.
    Still it took 40minutes to produce that file using an W3520@4GHz on an Adaptec 3805 hosting a 3R5 volume, there is no way that the SF controller is able to achieve that sort of compression (on the fly) as 40GB is written at a rate of ~100MB/s using a 60GB SF1 drive. (based on steady-state)
    I'll do some more tests when I get a few more of the items off my to-do list.
    -
    Hardware:

  18. #618
    Xtreme Enthusiast
    Join Date
    Jun 2011
    Location
    Norway
    Posts
    609
    I have filled up my M4 with os, apps and some music. A total of 42220109824 byte.
    M4 filled before testing.PNG

    Last thing before we start tomorrow is configuration of Anvils app.
    1: AMD FX-8150-Sabertooth 990FX-8GB Corsair XMS3-C300 256GB-Gainward GTX 570-HX-750
    2: Phenom II X6 1100T-Asus M4A89TD Pro/usb3-8GB Corsair Dominator-Gainward GTX 460SE/-X25-V 40GB-(Crucial m4 64GB /Intel X25-M G1 80GB/X25-E 64GB/Mtron 7025/Vertex 1 donated to endurance testing)
    3: Asus U31JG - X25-M G2 160GB

  19. #619
    Banned
    Join Date
    Jan 2010
    Location
    Las Vegas
    Posts
    936
    17.586 TiB written, 54 hours, sa177: 72/72/1392

    I included screenshots again today, but on subsequent updates I will just post the numbers above. I'll also post smart attribute 178 if it ever changes.

    2 days:
    smart_20110630_1054.png
    SAMSUNG 470 Series SSD_1GB-20110630-1052.png
    Last edited by johnw; 06-30-2011 at 10:55 AM.

  20. #620
    Banned
    Join Date
    Jan 2010
    Location
    Las Vegas
    Posts
    936
    Quote Originally Posted by Vapor View Post
    First, I took two of my drives, C: and D:, which are comprised of OS and applications (C:) and documents (D:, .jpg, .png, .dng, .xlsx probably make up 95% of the data on it) and froze them into separate single-file, zero compression .rar documents.
    Nice job, and beautiful graphs!

    There are two more data points I'd be interested in seeing: the compression your Sandforce drive achieves on your C: and D: not-compressed archive files.

    I guess you can measure it by just observing the SMART values on the drive, then copy one file to the drive, then look at the SMART values again to find the compression (assuming that attribute for actual flash writes is accurate). Maybe you have to delete the file and re-copy it several times to get an accurate measurement?

  21. #621
    Admin Vapor's Avatar
    Join Date
    Feb 2005
    Location
    Ann Arbor, MI
    Posts
    13,107
    Quote Originally Posted by Anvil View Post
    @Vapor
    Superb job on collecting compression data and yes, it's based on 7Zip Fast compression ratio. (could be Fastest, will check)

    Looking at my tests on the SF2 controller it couldn't keep up with the ratio that 7Zip Fast(est) produces, not sure how the SF1 handles vs the SF2.

    For reference, I 7Zipped one of my VM's earlier today (Windows Server 2008 R2, SQL-Server, + some data) and it ended up being ~50% of the original size using 7Zip Fastest.
    Still it took 40minutes to produce that file using an W3520@4GHz on an Adaptec 3805 hosting a 3R5 volume, there is no way that the SF controller is able to achieve that sort of compression (on the fly) as 40GB is written at a rate of ~100MB/s using a 60GB SF1 drive. (based on steady-state)
    I'll do some more tests when I get a few more of the items off my to-do list.
    I have zero doubt hardware designed for compression/dedup could do twice (at least) what our CPUs do with just a 1W power envelope....but that doesn't mean the SF1 and SF2 controllers can do it. It's a safe bet they can't and their compression levels are weaker than the weakest RAR/7zip setting--too bad there's no way of running their compression levels on our CPUs to see what they can do with more precision than the 64GB (or 1GB SF2) resolution the SMART values give.

    Almost done with the charts of all the drives so far (minus the V2-40GB...not sure whether or not to include that as testing essentially errored-out). Including a new chart with normalized writes vs. wear, which is kind of necessary considering drives of different sizes are getting entered into testing; writes will be normalized to the amount of NAND on the drive, not the advertised size.

    Working on bar charts with writes from 100-to-0 wear as well as total writes done so far. 100-to-0 wear will be extrapolated until MWI = 0 and then frozen...so when MWI > 0, total writes will be less than 100-to-0 but after MWI hits 0, total writes will be greater than 100-to-0. Would "MWI Exhaustion" be a better name for the 100-to-0 bar?

  22. #622
    Xtreme Mentor Ao1's Avatar
    Join Date
    Feb 2009
    Posts
    2,597
    Whoa, that is still running fast. It took One_Hertz 9 days to write that much. (longer still for Anvil)

  23. #623
    Admin Vapor's Avatar
    Join Date
    Feb 2005
    Location
    Ann Arbor, MI
    Posts
    13,107
    Quote Originally Posted by B.A.T View Post
    I have filled up my M4 with os, apps and some music. A total of 42220109824 byte.
    M4 filled before testing.PNG

    Last thing before we start tomorrow is configuration of Anvils app.
    I say just default values like what johnw is doing

    12GB for the free space, 0-fill, allow dedup.

  24. #624
    Banned
    Join Date
    Jan 2010
    Location
    Las Vegas
    Posts
    936
    Quote Originally Posted by Ao1 View Post
    Whoa, that is still running fast. It took One_Hertz 9 days to write that much. (longer still for Anvil)
    Yup, the Samsung 470 can sprint, but let's see if it has the stamina...

  25. #625
    Admin Vapor's Avatar
    Join Date
    Feb 2005
    Location
    Ann Arbor, MI
    Posts
    13,107
    Quote Originally Posted by johnw View Post
    Yup, the Samsung 470 can sprint, but let's see if it has the stamina...
    Not looking like it

    Jun30Host.png

    This isn't even the normalized writes chart where the Samsung (and C300 and M4 and later SF-1200s) will be 'corrected' for being larger.

Page 25 of 220 FirstFirst ... 15222324252627283575125 ... LastLast

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •