Slow Speeds on new SSD in eSATA enclosure

[NewToSSD]

I am using an OWC Mercury Elite Pro mini eSATA ( and USB 3.0) enclosure with a Crucial MX500 1TB SSD.

Both the eSATA and USB 3.0 transfer rates are not much faster than a 7200 RPM mechanical hard disk,
about 175 MB/s write and 225 MB/s read for eSATA.

The drive is rated at about 550 MB/s.

I am using a Gigabyte GA-P67X-UD3-B3 motherboard with regular BIOS Revision F8.

It is running Windows 7 Pro 64 bit.

The file system is NTFS with a Master Boot Record partition table.

The eSATA controller is an Intel P67 SATA 6 GB/s.

I am using IDE mode not AHCI because I have HDD's in the system that require IDE mode for compatibility.
I can power on or off those HDD's independently when the system is off and thereby choose which drive to have on,
but I need to leave the BIOS set to IDE mode to avoid having to go into the BIOS settings everytime I boot my system and change value back from AHCI to IDE if I am going to boot and use an HDD.
The slow speeds I observe are when using just the SSD alone with the controller set to IDE.

My USB 3.0 controller is Etron EJ168.

The USB 3.0 transfer rate is a little faster than the eSATA rate, about
195 MB/s write and 245 MB/s read for USB 3.0.

Are these speeds (both the eSATA connection and the USB 3.0 connection) normal for this enclosure given my settings,
or do I have a defective enclosure or defective SSD ???

Thank you for your help.

 

[slavie]

How are you testing your read/write speeds?
Should be faster than that. Would you be able to plug that drive directly into the MB and test read-write speed without the ext enclosure? That way you can confirm whether your MB and/or the Crucial drive is not the problem.
Make sure to connect to the SATA3 port on the mobo. That mobo has both native SATA3, SATA2, and extra controller SATA3. The native/chipset SATA3 would obviously be best.
You're never gonna get the same speed with the overhead of the ext encl, but there shouldn't be a 50% hit.

 

[drescherjm]

Both the eSATA and USB 3.0 transfer rates are not much faster than a 7200 RPM mechanical hard disk,

Should still be orders of magnitude faster in 4K random reads and writes than a 7200 RPM hard drive

 

[E4g1e]

There are several reasons for this slow performance:

1) The P67 chipset itself is very limited in the number of SATA III ports that it natively supports. Only two of the six SATA ports are capable of SATA III bandwidth; the other four are restricted to SATA II bandwidth. And it is very, very likely that the eSATA port is only capable of SATA II (not SATA III) bandwidth. Thus, you will never see more than about 270-ish MB/s out of your system's eSATA port.

2) Your motherboard also has a Marvell SATA III controller, which is also slower than the native Intel SATA III ports. Worse, the Marvell SATA III controller is somewhat buggy to begin with.

3) The P67 chipset does not support USB 3.0 natively. Instead, USB 3.0 support is added on using a third-party (in this case, Etron) controller on the motherboard. And Etron's USB 3.0 controller IMHO does not support UAS (USB-Attached SCSI); instead, it is permanently locked to the BOT (Bulk-Only Transport) mode like a typical USB 2.0 controller. Therefore, throughput from the Etron controller is limited to about half the bandwidth of what USB 3.0 itself is theoretically capable of.

4) Finally, Windows 7 itself does not natively support USB 3.0, requiring third-party drivers in order to even enable this mode at all. Unfortunately, like the Etron controller itself, the available Windows 7 drivers for the Etron controller are restricted to BOT-only transfers, which again restrict maximum throughput severely.

 

[scarletleopard]

I am using an OWC Mercury Elite Pro mini eSATA ( and USB 3.0) enclosure with a Crucial MX500 1TB SSD.

OP, did you ever figure this out? I have the USB-C version of the Elite Pro Mini and have been getting massively inconsistent performance on the eSATA port. In my case, I have none of the potential issues or bottlenecks suggested in other replies. Performance on USB is perfect. On eSATA it will benchmark normally one moment, then the next it will be at around SATA2 speeds. I don't understand how what seems to be a direct passthrough (the activity LED doesn't even flash in eSATA mode so they even bypassed that!) can be causing such fluctuations. I have tried several PCs and different eSATA cables to no avail.

 

[xx0xx]

OP, did you ever figure this out? I have the USB-C version of the Elite Pro Mini and have been getting massively inconsistent performance on the eSATA port. In my case, I have none of the potential issues or bottlenecks suggested in other replies. Performance on USB is perfect. On eSATA it will benchmark normally one moment, then the next it will be at around SATA2 speeds. I don't understand how what seems to be a direct passthrough (the activity LED doesn't even flash in eSATA mode so they even bypassed that!) can be causing such fluctuations. I have tried several PCs and different eSATA cables to no avail.

I don't know this for sure- but is it possible that eSATA ports in general are less consistent because of controllers, shared resource, etc- in other words- maybe more susceptible to fluctuations over time based on the system? For instance, I know that sometimes when certain regular SATA ports are used on motherboards, it can turn off or limit speeds of M.2 slots, or other SATA ports, etc because of the way they're set up to use resources. Maybe eSATA is affected more by this than USB?

Again, speculation on my part, but it is curious that you see no issue on USB, but always fluctuations on eSATA. I suppose the other possibility, since it happens on multiple PCs, which would have different boards/controllers on the PC side, is that the Elite Pro Mini's eSATA interface/controller itself is subpar? A lot of factors to consider...

 

[NewToSSD]

OP, did you ever figure this out? I have the USB-C version of the Elite Pro Mini and have been getting massively inconsistent performance on the eSATA port. In my case, I have none of the potential issues or bottlenecks suggested in other replies. Performance on USB is perfect. On eSATA it will benchmark normally one moment, then the next it will be at around SATA2 speeds. I don't understand how what seems to be a direct passthrough (the activity LED doesn't even flash in eSATA mode so they even bypassed that!) can be causing such fluctuations. I have tried several PCs and different eSATA cables to no avail.

Many of the points made by people responding to my post ultimately turned out to be correct in part or whole.

I contacted OWC and asked for another unit to try and they graciously sent me one.
The results from testing with it proved to be identical. (Ultimately, I kept both enclosures.)
This made me carefully re-examine my setup. This was illuminating.

The first thing I discovered was what I thought was a SATA III connection was actually a SATA II connection. The sata cable to the eSATA jack and one adjacent were both black and intertwined and I mistakenly thought that I had the correct one. I did not. Pilot error. Arghhh!!!

Even though I was loath to start moving wiring harnesses around on a 9 year old motherboard, I steeled myself and connected to a real SATA III connection. I observed an immediate increase in speed over the SATA II and also over the Marvell controller speeds.

Months later I updated to Windows 10 and weeks later set my P67 controller to AHCI. In many but not all tests, that resulted in a further increase in speed. Also with Windows 10 I began using the Windows 10 USB 3.0 controller driver which gave a small bump to the speed there also.

My drives are still NTFS + MBR and it is possible that an upgrade to UEFI + GPT might get me closer to the advertised speeds, but I do not dare try that on a board so old. That will have to wait till a motherboard / cpu / memory upgrade someday.

And yes, I confirm that the activity light only blinks on USB connection, not eSATA. OWC did not even know that till I told them.

So the combination of my mistake, the old style setup on old hardware with an old OS all were the sources of my problems.

I hope some or all of this information is helpful to you.