Crucial 256 GB M4 Review

With prices falling, SSDs are rapidly becoming mainstream and are included in a vast number of new builds. The Crucial RealSSD M4 series of SSDs are the newest drives from Crucial, available in 64 GB, 128 GB, 256 GB and 512 GB models. Crucial have kindly supplied a 256 GB unit for review, along with a data transfer kit.

The Hardware

The drive itself came in an anti-static packet, as would be expected, and was accompanied by the data transfer cable and disc. Two quick install guides were also included – one for installing the SSD itself, and one for the data transfer kit. This product is currently priced (in the UK) at approximately £300 (without data transfer kit), making it around £50 cheaper than the OCZ Vertex 3 that we have reviewed previously. Across the pond prices come in at approximately $370 for the M4, versus $440 for the Vertex 3. This places the M4 towards the lower end of the price range for SATA 3 SSDs.

The solid state drive itself and the enclosed kit
The solid state drive itself and the enclosed kit

Crucial claims, however, that this drive is capable of some serious performance. Read speeds of 500 MB/s and write speeds of 260 MB/s are claimed, along with random 4k read speeds of 50K IOPS and a MTBF (mean time between failure) of 1.2 million hours. This is all topped off with a 3 year warranty, which seems to be about the standard length for this technology. Naturally, this warranty becomes far shorter – zero, in fact – if you open it up to have a look at the innards, so we have an image here for you.

The SSD before I had my screwdriver in hand ...
The SSD before I had my screwdriver in hand ...

The unit came with the 0009 firmware
The unit came with the 0009 firmware

The screwdriver target...
The screwdriver target...

The PCB inside.
The PCB inside.

 

The memory chips are, rather unsurprisingly, Micron branded. It can be seen from the photograph that the controller is a Marvell 9174 chip, unlike the slew of Sandforce-based units on the market.

The Data Transfer Kit

While I’m always keen on a clean Windows reinstall, I gave the included data transfer kit a go. There are instructions (on paper) provided, but the video guide from Crucial is more informative and easier to follow, not that it’s a particularly complex procedure. The process was painless; I plugged the SSD into the suplied USB gizmo before booting from the CD. Copying across the (admittedly not very full) Windows 7 install was quite quick and worked perfectly. This may well save some people a considerable amount of time, but for the rest of the tests here I used a clean Windows 7  installation. The data transfer kit is likely more useful for people who would rather not spend the time reinstalling Windows, or for people who would like to upgrade to an SSD in the most straightforward way possible.

Using the enclosed transfer kit to copy a working Windows 7 installation onto the SSD
Using the enclosed transfer kit to copy a working Windows 7 installation onto the SSD

 

Test Set-up

The following hardware was used to test the  SSD; all benchmarks were run at least twice, and scores are reported as the average of the runs.

  • AMD A6-3650 APU @ 2.6 GHz (stock)
  • Biostar A75MH motherboard; the A75 chipset includes the necessary SATA3 (6 Gbps) support to get the most out of this unit
  • 4 GB (2x 2GB) Kingston RAM
  • BFG MaxCore GTX 260 Video Card
  • Antec CP-1000 PSU
  • LiteOn BD-ROM
  • Windows 7 Professional (64 bit)
  • Seagate Barracuda 7200.10 320 GB HDD or Crucial 256 GB M4 SSD

Benchmarking

Without any tweaking, the SSD made a major difference to the PCMark7 score right off the bat, raising it from 2147 to 3434. This is not surprising, compared to an older hard drive, but the difference it made to the overall score for the system surprised me quite a bit. Running other benchmarks on the SSD revealed excellent read and write speeds, often in excess of 400 MB/s for reads and 260 MB/s for writes. However, a further improvement is obtained by switching to AHCI mode in the BIOS. If transferring an older Windows install across, attempting to switch from IDE to AHCI causes Windows to flat out refuse to boot, while a clean install works fine. Re-running PCMark7 after reinstalling Windows 7 under AHCI mode increased the score slightly to 3506.
A series of other benchmarks were run to ascertain the performance of the M4 (which all showed best performance in AHCI mode). ATTO and AS SSD are useful ‘click and go’ benchmarks. ATTO shows read speeds of up to 550 MB/s and write speeds of up to 275 MB/s. AS-SSD records slightly lower, yet similar, speeds and gives the drive scores of 251/300/676, versus 260/345/734 for the Sandforce SF2281-based OCZ Vertex 3. While the Crucial drive can keep up in terms of read speeds, the access times are typically 1.5 – 2x that of the OCZ drive, and it falls behind in terms of write speeds.
CrystalDiskMark throws up similar numbers, with up to 500 MB/s reads and 270 MB/s writes; write speeds approach the Vertex 3 when random data is used as opposed to 0Fill and 1Fill, perhaps due to the data compression that renders Sandforce controllers among the fastest on the market today.
IOMeter was also employed to test this drive. All Overclockers.com reviewers use the same set-up file to ensure that results are comparable between reviews. Four tests were conducted: 2 MB sequential reads, 2 MB sequential writes, 4 K random reads and 4 K random writes. The results are tabulated below.
TestIOPSMB /sI/O Response TimeCPU Utilisation
AverageMaximum
2 MB Sequential Reads263526243 ms249 ms2.46%
263527243 ms249 ms2.73%
2 MB Sequential Writes130259493 ms517 ms2.10%
128256499 ms530 ms2.21%
4 K Random Reads17299680.230 ms7.198 ms19.81%
17447680.228 ms7.494 ms20.43%
4 K Random Writes 577212250.554 ms6.032 ms30.73%
577292260.554 ms9.050 ms30.65%

Finally, some comparisons between the OCZ Vertex 3 and the Crucial M4, presented in graph form. These essentially summarise what has been said above; the M4 keeps up with the Vertex 3 in read speeds, but loses out when it comes to writes. This is most likely a consequence of the controller. Notably, the M4 shows often much better read speeds for the smaller transfer sizes.

Most results are plotted in megabytes per second, although some are plotted in terms of relative performance versus the M4.

 

Conclusions

While the M4 doesn’t compete with the OCZ Vertex 3 in terms of write speed, its read performance is on par. Given the reliability issues that some drives built around Sandforce controllers are experiencing, the M4 looks like a more robust alternative that still delivers excellent performance. It lives up to, and in some cases exceeds, the claimed read and write speeds. In addition, the quite attractive pricing in the US makes this one to consider carefully when SSD-shopping. While other sizes are also available, remember that performance is not necessarily the same, and would be expected to be slightly lower for the smaller drives.
For these reasons, I’m happy to label this as  ‘Overclockers Approved’.
About David 21 Articles
I am an Overclockers.com Editor and OCForums Super Moderator. I help prepare articles for publication, as well as writing the odd article, news piece or review myself. I'm very keen on maintaining a high quality of writing in the articles on Overclockers.com, and emulating the objectivity and clear writing style of scientific journals. If you're writing for Overclockers.com and need some help with preparing your piece, or presenting data, please feel free to email me.I'm a chemist by trade, having completed a degree at the University of Edinburgh, Scotland in 2008. I'm currently studying towards a PhD at the University of Strathclyde (Glasgow, Scotland) in Physical Organic Chemistry on the topic of substrate structure/activity relationships in ruthenium-alkylidene catalysed olefin ring-closing metathesis. In my spare time I like to relax with a book or browse the forums.

Loading new replies...

Avatar of Janus67
Janus67

Benching Team Leader

17,200 messages 518 likes

Nice review, David.

I don't think I saw it mentioned, but I believe that the [write] performance gains from the Sandforce 22** series really show up when using compressible data vs non-compressible, and with non-compressible they are about the same. Do you know if/which of the different benchmarks use which type of data?

Reply Like

Avatar of Sentential
Sentential

Contributing Member

14,206 messages 0 likes

Any word of long-term reliability? Only reason I didnt buy an SF drive is how poorly my agility II did and since returned. In spite of the obvious performance benefits I went with a pair of intel drives instead.

Reply Like

Avatar of deathman20
deathman20

High Speed Premium Senior

15,991 messages 10 likes

Got to ask is the transfer cable that comes with it now just USB 2.0 or USB 3.0?

It does come in handy if you need a fast portable drive (SSD that is) to carry around if it comes with a USB 3.0 interface. I know the one I got sadly was just a USB 2.0 interface a while back with my 128Gig drive. Still nice for those situations where you need an adaptor though wish it was of the faster type that is out there.

Reply Like

S
Storageguy2000

New Member

1 messages 0 likes

Nice review, but something looks wrong on the 4K Iometer graph

David,

Nice summary on the M4. Great work translating all the test runs to the visual graphs for comparison. It makes looking at your two separate reviews easier to compare.

Do you have an error in your final graph that shows the results for the Iometer 4K runs? It says sequential reads and writes, but I think the 4K tests are random. Therefore I assume the graph shows the random reads are significantly lower for the M4 compared to the OCZ Vertex 3. Am I somehow reading that wrong?

Reply Like

Avatar of David
David

Forums Super Moderator

15,810 messages 10 likes

Thanks for the comments guys . Just a quick reply for now as I'm on a phone.
I'll check which benchmarks use compressable data - I suspect those where the Vertex is clearly better.
I think the cable is usb2 unfortunately, but I'll check when i get home. My desktop has USB 3.
Not sure about reliability from first hand experience - only had it a few weeks. I've heard no complaints thus far, lookking on the web.

The graph axis has a mistake - it's 4k random. I'll fix it tonight or tomorrow .

Cheers,
D

Reply Like

Avatar of EarthDog
EarthDog

Gulper Nozzle Co-Owner

76,334 messages 3,088 likes

Excellent review David!

Im not sure off the top of my head but my V3 240GB review does mention which is which I believe.

Reply Like

Avatar of krag
krag

Classifieds Moderator

6,537 messages 20 likes

Very nice review David. I take it that the OCZ Veritex has the Indilix controller? Which is faster?

Reply Like

Avatar of Janus67
Janus67

Benching Team Leader

17,200 messages 518 likes

I believe only the new OCZ (Octane) drives have the Indilix controller

Reply Like

Avatar of BossBorot
BossBorot

Member

2,936 messages 0 likes

Vertex 3, Vertex 3 Max IOPS, and Agility 3 are all sandforce 2281 with the difference between them being the type of NAND used. They use sync, toggle, and async in the order previously mentioned.

The new octane used the first new indilix controller to see retail in something like three years.

Any word of long-term reliability? Only reason I didnt buy an SF drive is how poorly my agility II did and since returned. In spite of the obvious performance benefits I went with a pair of intel drives instead.

Many buy the M4 specifically due to its reliability vs the faster sandforce 2281 drives that have a few issues.

Reply Like

Avatar of krag
krag

Classifieds Moderator

6,537 messages 20 likes

Excellent and useful information in the featured article and this thread. Thanks guys. :)

Reply Like