Plextor is one of those OEMs who have quietly been making their way into the SSD market. They don't have an aggressive marketing engine like for example OCZ does. Their drives are not featured on NewEgg's front page or advertised on most technology sites. If you have read our Plextor M3 review, you might remember that I had not even heard of Plextor before they contacted me. Older users may remember the name Plextor from times when optical drives were relevant, but for years Plextor was out of the spotlight because optical drive performance stopped being relevant for the majority of people and Plextor didn't really have any other products. In 2010, Plextor's ship took a new destination and that was the SSD market.

SSD prices have been dropping significantly over the last couple of years. Especially in the last few months, there have been lots of discounts. For example, the 128GB Vertex 4 had suggested retail price of $179 when it was released in April. Right now it's selling for $120 at NewEgg. That's 33% reduction in price in less than four months. Even more extreme example would be Crucial's m4. When it was originally released in April 2011, the 512GB version had a suggested retail price of $1000. Currently the exactly same drive retails for $400 at NewEgg, and there have been sales bringing the price down as low as ~$350.

The drop in prices has also increased consumer interest in SSDs. You no longer need to spend half of your entire PC budget on an SSD large enough to hold more than just your OS. With more buyers looking for SSDs, there is room for more manufacturers as well. SandForce's licensing strategy has allowed pretty much any hardware company to enter the SSD market, the most recent entrant being MSI. While SandForce SSDs are good in performance and are usually competitively priced, they are all more or less the same (Intel's custom firmware enabled SF drives being the exception). In the end, there are very few SSD OEMs that have truly unique SSDs. Unique in this context means that you at least have your own firmware. Intel, Samsung, Micron/Crucial, OCZ, Toshiba, SanDisk and of course Plextor are probably the most known manufacturers with their own firmware and even controller in some cases. These OEMs also happen to be the largest in the channel SSD market, which is no coincidence. 

While Plextor is still far away from gaining Intel or OCZ status in the SSD world, they are on the right path. We were very pleased with Plextor's M3 and M3 Pro when we reviewed them. Performance was great and both drives were backed by a 5-year warranty (more on reliability in a bit). The only real complaint we had was about pricing, which was not necessarily enough competitive to keep up with the constant price drops. Of course, there were sales that brought the M3's price down to the level of other SSDs, but in most cases you still had to pay premium if you wanted a Plextor SSD.

The M5S that we'll be looking at today is all about cutting costs while still providing the same performance and (hopefully) reliability that the M3 and M3 Pro provided. Without further delay, lets start off with a specification table:

Plextor M5S Specifications
Model	PX-64M5S	PX-128M5S	PX-256M5S
Raw NAND Capacity	64GiB	128GiB	256GiB
Usable Capacity	59.6GiB	119.2GiB	238.5GiB
Number of NAND Packages	8	16	16
Number of Die per Package	1	1	2
NAND	Micron 25nm synchronous MLC NAND
Controller	Marvell 88SS9174-BLD2
Cache	128MB DDR3	256MB DDR3	512MB DDR3
Sequential Read	520MB/s	520MB/s	520MB/s
Sequential Write	90MB/s	200MB/s	390MB/s
4K Random Read	61K IOPS	71K IOPS	73K IOPS
4K Random Write	25K IOPS	51K IOPS	70K IOPS
Warranty	3 years
MSRP	$100	$160	$300

The 128GB and 256GB models are nearly identical to the M3 in terms of performance. There are some minor changes but the only notable one is a 30MB/s increase in sequential write speed for the 256GB model. The 64GB model, on the other hand, has gone through some serious performance reshuffling: sequential write has dropped from 175MB/s to 90MB/s and 4K random write from 40K IOPS to 25K IOPS. 

Comparison of NAND Interfaces
	ONFi				Toggle-Mode
Manufacturers	IMFT (Intel, Micron, Spectec), Hynix				Toshiba/SanDisk, Samsung
Version	1.0	2.0	2.x	3.0	1.0	2.0
Max Bandwidth	50MB/s	133MB/s	200MB/s	400MB/s	166MB/s	400MB/s

Plextor couldn't tell us why they switched NAND suppliers for the M5S but I believe it has to do with price and supply. Remember that the M5S is all about cutting costs. Often times compromises have to be made in order to cut costs sufficiently. In this case, the 64GB model has compromised performance while the bigger capacities continue to run at roughly the same speeds.

A Word on Reliability

The M3 and M3 Pro came with 5-year warranties but the M5S only comes with 3-year warranty. The quick conclusion would be that the M5S must be inferior in some way because Plextor would not give it the same 5-year warranty as before. Plextor is obviously claiming that the change in warranty is only to differentiate their entry-level M5S and a future performance model. I don't doubt that and I even wanted Plextor to provide some differentiation between their SSDs because the M3 and M3 Pro were too similar. Offering different warranties is one good way to do that.

Plextor says that all their SSDs go through the same validation process, regardless of the series. I don't know the exact specifics of their testing methods, but according to their website all SSDs are tested for 20 hours in a high temperature burn-in test. Plextor is also claiming that their average annual failure rate is 0.5%. That's actually believable because according to third party data, Intel has had return rate as low as 0.1% but the 8MB bug increased the rate of returns. I checked NewEgg reviews for Plextor's M3 and M3 Pro and only 4.2% of the reviews (189 user reviews in total) were one or two eggs, which usually indicates a serious problem with the drive. That figure is roughly on-par with Intel's and Samsung's. I didn't calculate the exact figures for them but a quick look at NewEgg shows that their drives have around 3-7% of one or two-egg reviews. Of course, NewEgg user reviews are not the most reliable and the sample is also too small, but they give us some insight of reliability.

Plextor M3

I've been using the 256GB M3 as my boot drive since the review went live and I haven't had a single issue with the drive. One man's experience is of course not enough to declare that a drive is reliable, but I think it's safe to say that at least the M3 and M3 Pro are not plagued with issues. Assuming that the M5S follows the same path, there should be nothing to worry about.

Inside the M5S

Since the M5S does not include any software discs or 2.5" to 3.5" bracket, the packaging is noticeably smaller than M3's.

There are absolutely no add-ons included. The only contents of the package are the SSD itself and a quick installation guide. It should be noted that not even mounting screws are included, so hopefully your case came with a few.

The chassis at least looks like it's the same as in the M3. The color is a match and both measure in at 9.5mm in height.

The innards have changed quite significantly. The M3 and M3 Pro both had separate thermal pads for each main component (controller, NAND, DRAM), but the M5S only has one thermal pad which is for the controller. This is without a doubt a cost cutting measure. Thermal pads are not really necessary as SSDs don't generate much heat anyway.

The M5S is powered by Marvell's 88SS9174-BLD2 controller, just like the M3 and M3 Pro were too. I was hoping to see Marvell's 88SS9187, but perhaps Plextor is saving that for "M5 Pro". According to Plextor, the M5S does come with a different firmware than the M3 and M3 Pro, although I'm guessing that the M5S firmware was built upon the M3 (Pro) firmware. It's likely that the M5S firmware just adds Micron NAND support.

The only change in hardware appears to be in the NAND department. The M3 and M3 Pro used Toshiba's 24nm Toggle-Mode 2.0 MLC NAND, whereas the M5S is using 25nm ONFi 2.x MLC NAND from Micron. The change in NAND supplier has also resulted in changes in the PCB layout. There are now sixteen NAND packages, eight on each side of the PCB. The M3 and M3 Pro both had only eight NAND packages, regardless of the capacity. The 256GB model we have uses 16GB packages, each consisting of two 8GB dies.

Plextor stuck with Nanya as its DRAM supplier. There are two 256MiB DDR3-1333 SDRAM chips, giving the M5S a total of 512MiB of cache. 

The markings on the PCB actually say "M3S". Our unit is a pre-production sample, so that could be the reason but essentially the M5S is just an M3 with ONFi NAND.

Test System

CPU	Intel Core i5-2500K running at 3.3GHz (Turbo and EIST enabled)
Motherboard	AsRock Z68 Pro3
Chipset	Intel Z68
Chipset Drivers	Intel 9.1.1.1015 + Intel RST 10.2
Memory	G.Skill RipjawsX DDR3-1600 2 x 4GB (9-9-9-24)
Video Card	XFX AMD Radeon HD 6850 XXX (800MHz core clock; 4.2GHz GDDR5 effective)
Video Drivers	AMD Catalyst 10.1
Desktop Resolution	1920 x 1080
OS	Windows 7 x64

 

Random Read/Write Speed

The four corners of SSD performance are as follows: random read, random write, sequential read and sequential write speed. Random accesses are generally small in size, while sequential accesses tend to be larger and thus we have the four Iometer tests we use in all of our reviews. Our first test writes 4KB in a completely random pattern over an 8GB space of the drive to simulate the sort of random access that you'd see on an OS drive (even this is more stressful than a normal desktop user would see).

We perform three concurrent IOs and run the test for 3 minutes. The results reported are in average MB/s over the entire time. We use both standard pseudo randomly generated data for each write as well as fully random data to show you both the maximum and minimum performance offered by SandForce based drives in these tests. The average performance of SF drives will likely be somewhere in between the two values for each drive you see in the graphs. For an understanding of why this matters, read our original SandForce article.

Random read and write performance has not changed at all from previous generation. This was expected because Plextor was not claiming increased random read/write performance, and the performance figures from Plextor are almost identical; The 256GB M3 Pro has 75K IOPS random read and 68K IOPS random write. The respective figures for the 256GB M5S are 73K IOPS read and 70K IOPS write.  

 

Sequential Read/Write Speed

To measure sequential performance we ran a one minute long 128KB sequential test over the entire span of the drive at a queue depth of 1. The results reported are in average MB/s over the entire test length.

Sequential read and write performance has not changed dramatically either. Sequential read performance is right in-between the M3 and M3 Pro while sequential write performance is similar to the M3 Pro. As we discovered in our M3 Pro review, the standard M3 actually performed better in sequential write test with compressible data, although the M3 Pro was much faster when tested with incompressible data. 

AS-SSD Incompressible Sequential Performance

The AS-SSD sequential benchmark uses incompressible data for all of its transfers. The result is a pretty big reduction in sequential write speed on SandForce based controllers, while other drives continue to work at roughly the same speed as with compressible data.

Incompressible sequential read performance is in fact slightly (2-3%) faster than in M3 and M3 Pro. 

Incompressible sequential write speed falls right between the M3 and M3 Pro. Plextor is claiming up to 390MB/s sequential write on the M5S and it does get very close to that with incompressible data.

Performance vs. Transfer Size

All of our Iometer sequential tests happen at a queue depth of 1, which is indicative of a light desktop workload. It isn't too far fetched to see much higher queue depths on the desktop. The performance of these SSDs also greatly varies based on the size of the transfer. For this next test we turn to ATTO and run a sequential write over a 2GB span of LBAs at a queue depth of 4 and varying the size of the transfers.

Click for original size

Read performance is once again equivalent to the M3, regardless of the transfer size. Overall the read performance is excellent, only Crucial m4 and Samsung SSD 830 are faster at some transfer sizes.

Write performance could be better at small transfer sizes. As you can see in the graph, peformance at transfer sizes between 2KB and 64KB is noticeably slower than what Intel 520, Samsung 830 and OCZ Vertex 4 provide. 

Plextor is one of those OEMs who have quietly been making their way into the SSD market. They haven't been aggressive with marketing. Their drives are not featured on NewEgg's front page or advertised on most technology sites. 

Prominence in marketing, or a lack thereof, doesn't mean Plextor's drives are something to be ashamed of. We were quite pleased with the M3 and M3 Pro SSDs, and today we are looking at their new M5S SSD. While it's still equipped with the same Marvell 88SS9174 controller, Plextor has switched from Toggle-Mode NAND to ONFi NAND. The M5S is also using a brand new firmware which promises some improvements. So, to find out how the M5S performs in our tests, read on for the full review!

 

AnandTech Storage Bench 2011, Light Workload

Our new light workload actually has more write operations than read operations. The split is as follows: 372,630 reads and 459,709 writes. The relatively close read/write ratio does better mimic a typical light workload (although even lighter workloads would be far more read centric). The I/O breakdown is similar to the heavy workload at small IOs, however you'll notice that there are far fewer large IO transfers:

AnandTech Storage Bench 2011—Light Workload IO Breakdown
IO Size	% of Total
4KB	27%
16KB	8%
32KB	6%
64KB	5%

In our Light suite, the difference between the M5S and M3 is even smaller.

Performance Over Time & TRIM

Plextor's M3 and M3 Pro performed well in our torture testing. After 20 minutes of torture their performance was still bearable, which is what matters for consumer workloads (although even that is very extreme). Plextor relies on idle time garbage collection, so the performance restores when the drive is doing nothing. Plextor has always promoted their "True Speed" technology and I can see why. In fact, dirty state performance is one of the key features of the M5S as shown on Plextor's website.

I was excited to see if M5S brought even better garbage colletion as Plextor is advertising it so heavily. To begin testing, I ran HD Tach on a secure erased drive to get the baseline performance:

Next I tortured the drive by filling it with sequential data and then exposed it to 20 minutes of 4KB random writes (QD=32, LBA 100%):

Write performance drops to as low as 50MB/s for the first LBAs but the average write speed is up by 10MB/s compared to the M3. 

I let the drive sit idle for 30 minutes and reran HD Tach:

Performance is over 90% of clean state performance, which is very good. With more idle time and sequential writes, performance shhould get even closer to clean state numbers. 

Since 20 minutes of torture is not enough to put the M5S in its worst possible state, I secure erased the drive, filled it with sequential data and ran our torture test for 60 minutes:

And performance drops significantly as expected. 

I again let the drive idle for 30 minutes and reran HD Tach after that:

This is pleasant news. The M5S came from worst state to over 70% of clean state performance in only 30 minutes. For comparison, the M3 Pro only restored to 46% of clean state performance and it was idling for nearly two hours. 

Finally, I secure erased the drive one more time, filled it with sequential data and tortured for 60 minutes. After that, I formatted the drive in Disk Management to see if TRIM works properly (and it does):

 

Write Amplification

Write amplification has been reduced dramatically. The M3 Pro had relatively high worst case write amplification, although it was still acceptable. The M5S takes write amplification down to the level of most other non-SandForce drives. 

Conclusion

I'm pleased that Plextor has paid extra attention to the garbage collection in the M5S. The garbage collection in the M3 and M3 Pro was already good, but not perfect. The garbage collection in M5S is very aggressive if the drive is put into an extremely dirty state, which is good news because the most noticeable difference in performance comes when the drive is at its worst state.

I'm even happier about the fact that better garbage collection did not come at the expense of write amplification. In fact, the M5S almost halved estimated worst case write amplification compared to the M3 Pro. Aggressive garbage collection can come with serious downsides if not applied correctly. Plextor's approach has improved both garbage collection and write amplification, which is the optimal way when looking at the big picture.

Power Consumption

Idle power consumption is marginally lower compared to the M3 and M3 Pro. The difference is so small that it should not affect real world battery life, though.

Power consumption during sequential and random writes is noticeably higher than what the M3 and M3 Pro consumed. This is a small dissapointment because one of the main reasons I liked the M3 Pro was its power efficiency, yet high performance. I had hopes that the M5S would have adopted the power characteristics of M3 Pro, but that's definitely not the case. The change in NAND supplier might have some effect on power consumption, although not this big. Hopefully power consumption under load will be improved in future firmware updates - our review unit is using 1.00 firmware after all. 

 

Final Words

In a nutshell, the M5S performs exactly like the M3. It's not the fastest SSD in the market, but it provides balanced all-around performance. Some SSD manufacturers rely on compromises and only concentrate on certain areas of performance. A prime example is SandForce. It does well as long as you feed the drive with compressible data - once you switch to incompressible data it's a whole different story. Plextor's approach is to provide good performance regardless of the type or size of data, which I think is the best approach. 

While the overall performance matches the performance of the M3, there have been some welcome, and unwelcome, changes. The more aggressive garbage collection definitely helps if the drive is used in an OS without open TRIM support (*cough* OS X). However, most buyers will likely be running Windows with TRIM support, so the garbage collection is not a major selling point. Furthermore, the increase in load power consumption is a letdown. Desktop users have nothing to worry about, but for laptop owners it can be a big deal if you have a habit of running on battery most of the time. The power consumption is not awful but it was better in Plextor's previous generation SSDs, thus the disappointment.

NewEgg Price Comparison (7/16/2012)
	64GB	128GB	256GB	512GB
Plextor M5S	$100 (MSRP)	$160 (MSRP)	$300 (MSRP)	N/A
Plextor M3 Pro	N/A	$175	$300	$680
Plextor M3	N/A	$130	$250	$575
Corsair Performance Series Pro	N/A	$190	$330	N/A
Crucial m4	$65	$115	$210	$400
Intel 520 Series	$105	$150	$270	$520
Samsung 830 Series	$85	$150	$300	$720
OCZ Vertex 3	$100	$100	$190	$530
OCZ Vertex 4	$100	$120	$300	$700

As always, it all boils down to pricing at the end of the day. Plextor's press release says that the M5S will be available mid-July but I couldn't find it at any US resellers yet. Hence all we have is Plextor's suggested retail prices, which I wouldn't give much value. MSRPs tend to be higher than retail prices. A good example is Corsair's Force Series GS that was released a bit over a week ago. Its MSRPs were $190 for 180GB, $240 for 240GB, $350 for 360GB and $490 for 480GB; yet NewEgg was selling the drives for $175, $220, $320 and $450 on the launch day. I would expect the prices of M5S to drop to around the same level as the M3 is currently retailing for in a month or so. SSD prices fluctuate a lot anyway so it's impossible to recommend a drive based on pricing because the situation may be totally different tomorrow.

All in all, Plextor's M5S is a good drive but it doesn't really bring anything new to the market. The M3 has been available since late 2011 and the M5S is basically M3 with a couple changes. However, it's evident that something faster is in the works because why would Plextor dump the faster M3 Pro in favor of M5S. It seems that the limits of Marvell's 88SS9174 controller have already been reached, so it's probable that M5S' big brother will be based on Marvell's 88SS9187 controller. We don't know when that is going to be released, but given Plextor's ability to reach top of the class performance with the older 88SS9174, I'm eagerly looking forward to their "M5 Pro" and the controller not being the bottleneck anymore.