Jump to content


Photo

Samsung EVO SD Memory Card Review Discussion


  • You cannot start a new topic
  • Please log in to reply
5 replies to this topic

#1 Brian

Brian

    SR Admin

  • Admin
  • 5,207 posts

Posted 11 May 2014 - 07:06 PM

 

The Samsung EVO SD line offers great design, performance, and pricing, which results in a must buy for anyone looking for a quality SD card for their digital camera, DSLR, camcorder, or PC with an affordable price point.

Samsung EVO SD Memory Card Review

Brian

Publisher- StorageReview.com
Twitter - @StorageReview

 

#2 Sanpete

Sanpete

    Member

  • Member
  • 2 posts

Posted 20 July 2014 - 05:12 PM

Do you have any small-block (4K) random read/write results for this?


Edited by Sanpete, 20 July 2014 - 05:12 PM.

#3 pickles

pickles

    Member

  • Member
  • 2 posts

Posted 21 July 2014 - 07:01 PM

Samsung EVO 32GB SD -- The first time I ran CrystalDiskMark I got performance very similar to what was reported in the article: fast random writes, but mediocre sequential writes. This was a brand new card out of the package, never formatted or written before. 

-----------------------------------------------------------------------
CrystalDiskMark 3.0.3 x64 (C) 2007-2013 hiyohiyo
                           Crystal Dew World : http://crystalmark.info/
-----------------------------------------------------------------------
* MB/s = 1,000,000 byte/s [SATA/300 = 300,000,000 byte/s]


           Sequential Read :    46.325 MB/s
          Sequential Write :    19.457 MB/s
         Random Read 512KB :    41.880 MB/s
        Random Write 512KB :    35.283 MB/s
    Random Read 4KB (QD=1) :     4.623 MB/s [  1128.6 IOPS]
   Random Write 4KB (QD=1) :     1.084 MB/s [   264.8 IOPS]
   Random Read 4KB (QD=32) :     4.691 MB/s [  1145.2 IOPS]
  Random Write 4KB (QD=32) :     1.210 MB/s [   295.3 IOPS]


  Test : 1000 MB [I: 0.0% (0.0/29.3 GB)] (x5)
  Date : 2014/07/21 14:26:39
    OS : Windows 7 Professional SP1 [6.1 Build 7601] (x64)

When I ran another test the sequential writes improved but the random writes settled down.

 

-----------------------------------------------------------------------
CrystalDiskMark 3.0.3 x64 (C) 2007-2013 hiyohiyo
                           Crystal Dew World : http://crystalmark.info/
-----------------------------------------------------------------------
* MB/s = 1,000,000 byte/s [SATA/300 = 300,000,000 byte/s]


           Sequential Read :    46.465 MB/s
          Sequential Write :    27.873 MB/s
         Random Read 512KB :    42.091 MB/s
        Random Write 512KB :    23.261 MB/s
    Random Read 4KB (QD=1) :     4.596 MB/s [  1122.2 IOPS]
   Random Write 4KB (QD=1) :     1.668 MB/s [   407.3 IOPS]
   Random Read 4KB (QD=32) :     4.675 MB/s [  1141.3 IOPS]
  Random Write 4KB (QD=32) :     1.680 MB/s [   410.1 IOPS]


  Test : 1000 MB [I: 0.0% (0.0/29.3 GB)] (x5)
  Date : 2014/07/21 15:11:21
    OS : Windows 7 Professional SP1 [6.1 Build 7601] (x64)

 

I continued to run single tests, but the performance stayed about the same. I was curious about the initial behavior, so I did a full overwrite format using SD Formatter 4.0 to see if I could somehow mimic the as-new condition. It had no effect.

 

Question: Could the initial slow writes be caused by some sort of card initialization routine?

 

I was also seeing some poor results on single pass tests on the small random writes. Example:

 

  Random Write 4KB (QD=32) :     0.287 MB/s [    70.2 IOPS]
  Random Write 4KB (QD=32) :     0.305 MB/s [    74.4 IOPS]
 
As I understand it, CDM reports the best of the tests, so the 5X tests won't show the worst results.

Edited by pickles, 23 July 2014 - 12:39 PM.

#4 pickles

pickles

    Member

  • Member
  • 2 posts

Posted 23 July 2014 - 02:40 PM

Here are results on the Samsung EVO 32GB SDHC using ATTO Benchmark v2.47 (and SanDisk All-in-One reader):

 

AED3m4v.png

 

Looks pretty good, right? Let's run this test again (but this time using a Lexar SR1 reader):

 

4JL2JZV.png

 

That didn't go nearly as well. So let's try the SanDisk reader again:

 

7GQGted.png

 

Why so much slower now? Run the test again:

 

OUjZUel.png

 

And once again back on the Lexar SR1 reader:

 

Q8TYCKO.png

 

 

Why are write speeds so inconsistent? How was I able to get 35MB/s write in one test, then it suddenly drops on subsequent tests? Is the card controller trying to adjust?

 

If you are wondering about other possibilities that might explain this difference (computer, card readers, etc) , I've tested several other cards and none have behaved inconsistently except this Samsung EVO. I get consistent 90MB/s read and 85MB/s write on SanDisk Extreme Pro for example.


Edited by pickles, 23 July 2014 - 03:24 PM.

#5 continuum

continuum

    Mod

  • Mod
  • 3,540 posts

Posted 31 July 2014 - 12:58 AM

Tried a different benchmark?

 

Drive running out of clean LBA's and having to clean them up?

 

Heck, tried a few real-world benchmarks to see how it holds up?


#6 stomemouse

stomemouse

    Member

  • Member
  • 1 posts

Posted 03 August 2014 - 07:35 AM

So RND 4KB is on this card really slower than on my usual Samsung MicroSD 32GB class 6 ( B00BD3BFAM ) !?

 

           Sequential Read :    18.219 MB/s
          Sequential Write :    14.382 MB/s
         Random Read 512KB :    17.997 MB/s
        Random Write 512KB :    15.308 MB/s
    Random Read 4KB (QD=1) :     4.893 MB/s [  1194.6 IOPS]
   Random Write 4KB (QD=1) :     3.474 MB/s [   848.1 IOPS]
   Random Read 4KB (QD=32) :     4.978 MB/s [  1215.3 IOPS]
  Random Write 4KB (QD=32) :     3.505 MB/s [   855.6 IOPS]
  Test : 50 MB [F: 8.5% (2.5/29.3 GB)] (x1)
 
I wonder because I thought about buying one instead of the above for my RPi.

Edited by stomemouse, 03 August 2014 - 07:37 AM.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users