SM2NGFFMBU33.mainM.2 solid state drives are normally used in laptops, but they’re also small enough that you can also use them as really incredibly fast and spacious USB drives. Most USB drives use slower eMMC memory, so moving to a proper SSD provides a big step up in terms of transfer and access speeds. Startech.com have produced a M.2 enclosure, letting you take an M.2 SSD on the go. Let’s see how well it works!

 

Specifications

USB 3.0 interface with UASP supportAsmedia ASM1153 chipset
for M.2 (NGFF; B Key) drivesnot for PCI-E M.2 SSDs
Maximum data rate 5 Gbps (USB 3 / SATA III)USB powered
Black aluminium enclosure100 x 35 x 15 mm, 68 grams
Contains enclosure, USB 3.0 cable
SSD screw kit, mini screwdriver, manual
2 Year warranty
 

Design

IMG_3809

The enclosure has a pretty straightforward design; essentially a black aluminium tube, blank on one side and with the StarTech.com logo on the other.

On on end we have the USB 3.0 port, and a pair of screws to insert your M.2 SSD (which is not included).
SM2NGFFMBU33.B
IMG_3814
It’s worth noting that the enclosure requires the cable to operate, unlike most USB flash drives which have a USB tip. This means it’s a bit easier to plug the enclosure into crowded USB ports, but it also means you have to keep a USB 3.0 cable with you to use it.

 

Testing

Let’s get to the fun part – installing a drive and then seeing how fast it goes. This of course depends on the drive we put inside, so we’ll be using Samsung’s recent M.2 850 Evo drive.

Installation

OK, so installation. We have to take off that grille by unscrewing the tiny screws. I found the built-in screwdriver quite bad, so I used a (slightly larger) one of my own to do the job, eventually.

IMG_3815

OK, now it’s time to mount your M.2 SSD on the enclosure board.  There are a couple of different holes to accommodate different-sized M.2 SSDs; we’re using the largest size. The drive slides into the M.2 port, then is secured with the screw and nut. This felt really fiddly to me, and didn’t seem to be too well secured once finished. Still, once installed into the enclosure there’s no real opportunity for it to go anywhere.

IMG_3818

OK, so we’ve got the drive mounted now. Now, it’s just a case of sliding it back into the enclosure, and doing up the screws once again.

 

IMG_3816 IMG_3817

OK, we’re good. Let’s move onto some performance testing – using benchmarks and a a couple of real world tests too.

IMG_3810

Real world transfer time test

It took almost exactly 2 minutes to transfer 27.2 GB of large video files, at an average rate of 230 MB/s. Copying back was faster, taking just 1 minute and 40 seconds to copy the same amount at an average rate of 270 MB/s.

Screenshot 2015-03-26 00.05.01

By comparison, a standard USB 2.0 flash drive copies similar material at a speed of around 15 MB/s, while a standard USB 3.0 drive is around 40 MB/s – a fraction of the speed. A good internal SSD, by comparison, can hit 400-500+ MB/s.

Benchmarks

CrystalDiskMark

CrystalDiskMark has been one of my favourite benchmarks for evaluating storage speeds for a while now, thanks to its varied incompressible workload and extremely readable results. Here’s the 3.0.3 x64 version of the benchmark; all results are in MB/s.

Screenshot 2015-04-28 04.53.15

We see strong results here, besting the 300 MB/s barrier for sequential and 512K reads, and the 270 MB/s mark for their respective writes. This is about 60% of the speed of the full-size version of the Samsung 850 Evo.

CDM 3 ReadSeq512K4K4K QD32
Samsung 850 Evo51347238408
Samsung 840 Evo51547335397
Corsair Neutron GTX45037627328
Samsung 850 Evo (M.2 enclosure)3223161823
CDM 3 WriteSeq512K4K4K QD32
Samsung 850 Evo50447968352
Samsung 840 Evo50039086313
Corsair Neutron GTX48046869158
Samsung 850 Evo (M.2 enclosure)2752712231

AS SSD

AS SSD is another benchmarking tool quite similar to CrystalDiskMark, which uses predominantly incompressible data across a range of workloads including sequential tests, random performance and access times. First three numbers are MB/s and access time is in ms.

Screenshot 2015-04-28 05.23.49

 

Speeds are a little behind what was shown on CDM, but it’s still a reasonable result. Interestingly, the access times are around 3x slower than the internal 850 Evo.

AS SSD ReadSeq4K4K QD64Acc. TimeScore
Samsung 850 Evo516343820.059467
Samsung 840 Evo513323300.059413
Corsair Neutron GTX507253340.068N/A
Samsung 850 Evo (M.2 enclosure)30316210.18467
AS SSD WriteSeq4K4K QD64Acc. TimeScore
Samsung 850 Evo497662990.051415
Samsung 840 Evo497692070.054326
Corsair Neutron GTX473622950.062N/A
Samsung 850 Evo (M.2 enclosure)25621280.18175

HD Tune Pro 5.5

We conclude with HD Tune Pro, a benchmark which produces three scores for average read speed, read access time, and burst read rate.

Screenshot 2015-04-28 05.18.56

The burst rate here is again about 60% that of the internal Samsung 850 Evo drive. Once again, it’s quite reasonable performance for a USB 3.0 drive!

HD Tune Pro 5.50 ReadAverageAccess TimeBurst Rate
Samsung 850 Evo370 MB/s0.044 ms250 MB/s
Samsung 840 Evo358 MB/s0.041 ms231 MB/s
Corsair Neutron GTX340 MB/s0.059 ms231 MB/s
Samsung 850 Evo (M.2 enclosure)175 MB/s0.169 ms132 MB/s
 

Conclusion

Here’s what I’ve learned: the Samsung 850 Evo makes for an expensive but incredibly fast portable drive. The StarTech.com enclosure did well here; after a slightly fiddly installation I had no complaints. The enclosure felt solid and kept the drive well-ventilated, allowing its performance to shine through.

At £35, you can certainly get a reasonable amount for your money… but I still reckon they should have included a better screwdriver!

Pros

  • Quick way of turning an M.2 SSD into portable storage
  • Good performance
  • No heat issues

Cons

  • Quite fiddly to install
  • Screwdriver provided is not fit for purpose
  • M.2 SSDs aren’t cheap

Score

  • 9 / 10

Store link

About The Author
William Judd

Editor-in-Chief for XSReviews. Find me @wsjudd or on G+.

  • nster

    Could you try again with your own USB 3.0 Micro-B cable please? StarTech’s cables have issues

    • huh?

      • nster

        Many people have had issues with the cable provided with this, including less than optimal speeds and disconnects. I was wondering if the numbers you for were optimal speeds or were hindered by the cheap cable provided

        • Interesting. I don’t have the drive any more, so I can’t really test the difference between different cables. Still, thanks for your comment.

          • nster

            Thanks for responding so quickly :)