Super slow ssd sequential reading speed

limyueshing

Reputable
Feb 14, 2015
7
0
4,510
How can I fix this?

I am using an alienware x51 and I added a samsung evo 840 ssd for it. Since there doesn't exist an extra sata power port, I used a Y cable. Does that affect the performance?

-----------------------------------------------------------------------
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 : 50.123 MB/s
Sequential Write : 415.154 MB/s
Random Read 512KB : 173.300 MB/s
Random Write 512KB : 157.278 MB/s
Random Read 4KB (QD=1) : 36.797 MB/s [ 8983.7 IOPS]
Random Write 4KB (QD=1) : 115.773 MB/s [ 28265.0 IOPS]
Random Read 4KB (QD=32) : 362.923 MB/s [ 88604.2 IOPS]
Random Write 4KB (QD=32) : 150.695 MB/s [ 36790.7 IOPS]

Test : 2000 MB [F: 0.1% (0.1/100.0 GB)] (x5)
Date : 2015/02/15 18:55:16
OS : Windows 8.1 [6.3 Build 9600] (x64)
 
Solution
The power splitter should be sufficient. No worries there.

In your BIOS, is your SATA mode AHCI or IDE? If IDE, you should change to AHCI for best performance.

Are you using a clean install of Windows or did you clone a previous installation to the SSD (from a previous HDD for example)?
The power splitter should be sufficient. No worries there.

In your BIOS, is your SATA mode AHCI or IDE? If IDE, you should change to AHCI for best performance.

Are you using a clean install of Windows or did you clone a previous installation to the SSD (from a previous HDD for example)?
 
Solution


Not sure about the sata mode yet, I will check it later.

I've just done a clean re-install of windows 8 on my HDD and my SSD was used as the boot drive before. I used cloning to do that and also cloned the recovery partitions which I cannot format with disk management now. Will they affect reading speed?

 


Here's the description of my original hard drive in my receipt:
1TB (64MB Cache) 7200 RPM SATA 6Gb/s

I've confirm that my SSD also has a SATA transfer rate of 6Gb/s using Intel rapid storage, so I guess no?
 
Use Magician to enable "RAPID". Night and day difference on my SSD 850 EVO that I just installed. I have SATA II only and my PassMark results are now almost off the chart. (Seq.Write; 2602 MB/s.)

Never mind. Real world improvements nil. Apparently "Rapid Mode" fooled the test I was running.
 


From your original post:
"* MB/s = 1,000,000 byte/s [SATA/300 = 300,000,000 byte/s]"

That denotes a SATA II port.
Whatever drive is connected to it, be it HDD or SSD, will only run at SATA II speed.
For an HDD, it will not matter. An SSD will be significantly slower than normal.
 


Just checked Intel Rapid Storage, and my SSD has SATA transfer rate of 6Gb/s, that denotes a SATA III port right?

Even if its running at SATA II speed, isn't it too strange to have 50MB/s reading speed and 400MB/s writing speed?
 


Thats too unrealistic for me as their official page says its speed is ~500MB/s and doubles in rapid mode. What have u done?
 
I've just used Samsung's 840 evo restoration and magician, it appears that my SSD is working fine even not using AHCI. Here's its performance now with RAPID enabled:

-----------------------------------------------------------------------
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 : 7134.326 MB/s
Sequential Write : 2462.602 MB/s
Random Read 512KB : 6710.049 MB/s
Random Write 512KB : 3577.140 MB/s
Random Read 4KB (QD=1) : 950.709 MB/s [232106.7 IOPS]
Random Write 4KB (QD=1) : 320.774 MB/s [ 78314.1 IOPS]
Random Read 4KB (QD=32) : 865.157 MB/s [211219.9 IOPS]
Random Write 4KB (QD=32) : 266.132 MB/s [ 64973.5 IOPS]

Test : 2000 MB [F: 0.1% (0.1/99.5 GB)] (x5)
Date : 2015/02/16 0:18:59
OS : Windows 8.1 [6.3 Build 9600] (x64)

Thanks for your effort!
 

Yes, I'm a SSD newb. "Rapid" fooled the test I was using. I see no real improvements, and other test show none. My mistake.
 

TRENDING THREADS