discussion G. Skill Titan 256GB SSD Review

[AkG]

G. Skill Titan 256GB SSD Review​

Manufacture Page: SATA
Part Number: FM-25S2S-256GBT1
TechWiki Info: Titan - TechWiki
Price: Click Here to Compare Prices
Warranty: 2 year



As we have said in the past, if there is one constant in the computer industry it is change. While not all change is good, for the most part components are getting smaller, faster more efficient and even in some areas more durable. Take the new i7s for example; they run at temperatures which would have melted top of the line CPUs from just a few years ago. While the i7 is a good example of durability the absolute best example has to be the Solid State Drive revolution. These drives have no moving parts and can withstand g-forces which would literally shatter darn near any platter based hard drive.

While there is no denying SSDs are going to play at the very least a large part in the future of personal computing there have been some valid concerns raised about the earlier models. The two biggest of complaints are their price (both total cost and per GB) and their size. The first issue may still be relevant but hopefully, not for much longer. Just a year ago, a 32GB SSD would cost you an arm and a leg; today they can go for about the same price a normal laptop drive would cost (albeit a MUCH larger capacity 7200rpm laptop drive). Naturally, as the technology matures and more competition is born the demand for a product increases and prices fall like a proverbial rock. This is what is happening in the SSD industry today to the point where many SSDs are getting to the point where they are becoming a viable alternative to platter-based drives.

The second concern -regarding storage size- is also being alleviated. Where 32GB and 64GB were the norm (and priced in the reasonable range), 128GB and even 256GB drives are now starting to go mainstream. For most people 128GB is plenty of storage space for their laptops and even their desktops.

Unfortunately, there are some other concerns and these are deal breakers in our opinion; one of which is regarding durability. While an SSD can withstand huge heat and G-Force stress there has always been the nagging doubt about its performance longevity. This issue has been exacerbated by manufacturers selling SSDs with only a single year of warranty. The root of these concerns does not lie with SLC drives but with the more common MLC based ones. This is mainly a misconception based on a false perception; because SLCs have yet to fall in price as much as MLC and (thus are prohibitively expensive for the masses) they therefore must be "better". As in if it costs twice as much for half the amount of capacity it must be better…which we have shown in the past to be utter tripe and foolishness.

With recent advancements in manufacturing and "wear leveling" technology this concern really has not been valid for a while now. Yes MLC may not be as durable as SLC, but will in most cases last longer than your laptop! The fourth concern has to do with speed. It is only human nature to expect something which can cost a heck of a lot more than a normal drive to perform better as well. This too has been a weakness of MLC vs. SLC, as SLC is simply faster on a write cycle (one vs. two cycles). However, when you combine "big enough", "durable enough" and "cheap enough" with "fast enough" do you end up with a drive which is ready for Prime Time?

This question and those valid concerns are going to be the focus of today’s review, as it just so happens that we have in our possession a 256GB, MLC based, 2.5" Solid State Drive from G.Skill aptly named the Titan. This drive is supposed to be fast to the tune of 200MB/s read and 160MB/s write speed and weighing in at 256 GB, it has enough capacity for most consumers. While all those are great attributes for an enthusiast item, the largest selling point may be its retail price of around $680.

The Titan is starting to become available and you will soon be able to find it in many retailer and e-tailers throughout the country. If this is drive is as good as the hype surrounding it says it is, this will be the one of the few which many an enthusiast will be drooling (and fighting) over this year. So without further ado let’s put this puppy through its paces and see if the hype which surrounds this drive is well founded.

​

 

[AkG]

Specifications

Specifications

<img src="http://images.hardwarecanucks.com/image/akg/Storage/Gskill_Titan/gskill_titan_specs.jpg" border="0" alt="" />
​

 

[AkG]

Packaging and Accessories

Packaging and Accessories

​

In the past we were less than impressed with the packaging scheme G.Skill used to package and ship such an expensive item and they have certainly listened to our critique. This box may be small, though it is much thicker than the Patriot Warp v2 we reviewed recently, but it sure is a head turner. Just like an old school Jaguar XJ-S (or even more fittingly a new Tesla Roadster), this box exudes a presence which is almost palpable. It’s certainly is an aggressive color scheme but as the Titan IS a enthusiast model drive, a bit of machismo is fitting.

​

The box has all the necessary information about the drive itself, and while it may not have a picture of the Titan anywhere it’s really not needed. You either know what the Titan is or you really have no business owning it. It may be beautiful but it just fits this device and seems to enhance the mystic surrounding a RAID 0 Solid State Drive.

​

When you do open the box up and slip the internal package out, you will be greeted to one of the best, if not THE best packaging scheme we have seen in a while. Sure, it may only be high density foam hollowed out to make a form fitting package, but it is the shear elegance and neat factor which makes it great. In a nut shell you take this foam box and open the lid just like you would open a book. It works, it’s different and we instantly liked it.

​

Unfortunately, and just like the Patriot Warp v2 we recently reviewed, the list of accessories is for all intents and purposed non existent. You get a simple pamphlet and that is it. Where this is a cutting edge enthusiast piece of kit, a case badge or sticker would have been nice. Where is the right angle cable which came with the 64GB SLC we reviewed awhile back?!

All in all, we walked away extremely impressed with the Titan. It may not have any accessories worth talking about, but this seems to be the trend these days so we really can’t fault G.Skill that much for playing follow the leader. If the first impression of this drive proves to be a clear indicator of its performance, we are going to be in for one heck of a ride.

 

[AkG]

First Impressions

First Impressions

​

The all metal cover of the Titan is done is in an all black colour scheme with the only colour to be found anywhere being on the labels themselves. The “top” label is done in the same black fading to red design which graces the package the Titan came in. It’s a clean, no nonsense label which tells you what it is, who made and what size it is….and that’s it. No need for extraneous decals or goo-gaws here.

​

The back label is a study in contrast, as it is a fairly typical white with black lettering affair. The most notable features of this label (besides the fact it is skewed and was improperly applied…tsk, tsk) is the fact the Titan is rated to draw .8 of an amp off the 5v line. This means this drive will suck up to 4 watts of juice at most. This is a bit more than a 7200rpm laptop drive may use (usually in the 3 – 4 watt range during seeks) but it is still less than the X25’s full 1.0A rating.

​

One nice thing about the all black metal case is it contrasts nicely with the gold power and data port connectors.

​

Our first impressions of the Titan have done nothing to dissuade us from our initial positive impression we got from the packaging. While we are not enthused to see the skewed label, as this does point to lax QA, it is only a sticker. Now when we crack open this sucker and look inside, if we see anything not right, THEN we will be less enthusiastic about the Titan and its QA department. Until then, this drive is still shaping up to be one 100% kick arse drive.

 

[AkG]

Internal Impressions

Internal Impressions

Please remember that opening an SSD will effectively void your warranty.

​

To open the Titan all you have to do is remove four screws and gently pry the top lid off. Unlike the Patriot Warp v2 we recently reviewed this drive does not have any warranty void stickers in place BUT just because there are none, it doesn’t mean cracking open your shiny new Titan WON’T void your warranty. If you do open it up just assume you have nullified your warranty and pretend the stickers are of the invisible, printed on unobtanium variety.

​

The board layout is at first blush very similar to the Warp v2. At one end you have your SATA power and data connectors, with the I/O controller chips offset and then two rows of four NAND MLC flash chips. Meanwhile, the other side of this PCB has another 8 chips soldered to it (also in two rows of 4) making it for a grand total 16 NAND flash chips. However, unlike the v2 this G.Skill unit has not one but two I/O controller chips and a very small RAID controller chip located between the data / power ports and the two I/O controllers.

​

The I/O controller chips shown above are, as expected, JMicron model 602s. This is the same chip found in many SSDs out there but since this drive is internally set up as two SSDs on the one PCB there has to be two of them. Through the miracle of miniaturization the Titan really is two 128GB drives melded onto the one PCB and all crammed inside a standard 2.5” laptop form factor. That is one impressive feat!

​

The magic chip which allows this melding of two SSDs into one is the RAID controller chip shown above. This chip is also by JMicron; to be specific it is the JMB390 controller chip. This controller is a one-lane PCI Express to one-port Serial ATA II Host Controller which supports SATA II Gen2i and Gen2m, as well as Native Command Queuing (NCQ) and Hot Plugging. All in all it is a small, low power RAID controller which should make good fit for devices such as SSDs, assuming it is powerful enough that is. It is a shame that this controller does not seem to support external cache (or at least doesn’t have any in this setup). If it did then it really would have the potential of being the perfect mini hardware RAID controller for SSDs.

​

The NAND chips used in the Titan are Samsung K9MDG08U5M-PCB0. Once again we, using the online Samsung model decoder we can see these chips are 48 pin MLC DSP, 1st gen lead free (ROHS compliant), 2.7V ~ 3.6V, 50 nanosecod NAND chips which operate with Quad nCE (Quad Chip Enable control) & Quad R/nB (Quad Ready/Busy Output). This model is rated at a density of 128Gbits or 16GB per chip. Above the model number (and as stated earlier) we can see these were made in the 46th week of 2008 and below it we can see the batch number “FCIA87X2” (or at least what we assume is the batch number but is describe by Samsung vaguely as “Customer List Reference” only).

This certainly is an intriguing setup with top notch quality components. We are still not totally sold on the whole dual 602 controller setup, but it certainly is an interesting approach to the stutter problem. How effective will it be? Only time and testing will tell. Whether or not it is effective, we do have to wonder how much added thermal output this setup will create. After all, there are now 3 controller chips all creating heat compared to the more normal one. Let’s first look at whether or not it can walk the walk and not just talk the talk, then we will look at how hot it gets while doing it.

 

[AkG]

Testing Methodology

Testing Methodology

Testing a hard drive is not as simple as putting together a bunch of files, dragging them onto folder on the drive in Windows and using a stopwatch to time how long the transfer takes. Rather, there are factors such as read / write speed and data burst speed to take into account. There is also the SATA controller on your motherboard and how well it works with SSDs to think about as well. For best results you really need a dedicated hardware RAID controller w/ dedicated RAM for SSDs to shine. Unfortunately, most people do not have the time, inclination or monetary funds to do this. For this reason our testbed will be a more standard motherboard with no mods or high end gear added to it. This is to help replicate what you the end user’s experience will be like.

Even when the hardware issues are taken care of the software itself will have a negative or positive impact on the results. As with the hardware end of things, to obtain the absolute best results you do need to tweak your OS setup; however, just like with the hardware solution most people are not going to do this. For this reason our standard OS setup is used. However, except for the XP load test times we have done our best to eliminate this issue by having the drive tested as a secondary drive. With the main drive being a WD 320 single platter drive.

For these tests we used a combination of the ATTO Disk Benchmark, HDTach, HDTune, Cystal Disk Benchmark, h2benchw, SIS Sandra Removable Storage benchmark, and IOMeter for synthetic benchmarks.

For real world benchmarks we timed how long XP startup took, Adobe CS3 (w/ enormous amounts of custom brushes installed) took, how long a single 4GB rar file took to copy to and then from the hard drives, then copy to itself. We also used 1gb of small files (from 1kb to 20MB) with a total 2108 files in 49 subfolders.

For the temperature testing, readings are taken directly from the hottest part of the drive case using a Digital Infrared Thermometer. The infrared thermometer used has a 9 to 1 ratio, meaning that at 9cm it takes it reading from a 1 square cm. To obtain the numbers used in this review the thermometer was held approximately 3cm away from the heatsink and only the hottest number obtained was used.


Please note to reduce variables the same XP OS image was used for all the hard drives.

For all testing a Gigabyte PA35-DS4 motherboard was used. The ICH9 controller on said motherboard was used.

All tests were run 4 times and average results are represented.

Processor: Q6600 @ 2.4 GHZ
Motherboard: Gigabyte p35 DS4
Memory: 4GB G.Skill PC2-6400
Graphics card: Asus 8800GT TOP
Hard Drive: 1x WD 320
Power Supply: Seasonic S12 600W Performance Testing

 

[AkG]

Read Performance

Read Performance

For this benchmark, HDTach was used. It shows the potential read speed which you are likely to experience with these hard drives. The long test was run to give a slightly more accurate picture.

We don’t put much stock in Burst speed readings and this goes double for SSD based hard drive. The main reason we include it is to show what under perfect conditions a given drive is capable of; but the more important number is the Average Speed number. This number will tell you what to expect from a given drive in normal, day to day operations. The higher the average the faster your entire system will seem.

​

Well one thing is for certain: this bad boy has power to spare when it comes to read speeds. The average speed is simply phenomenal and the burst is what you would expect from darn near any high end drive out there. If someone had said even just two years ago we would see single drive performance numbers of 166/225+ MB/s we would have laughed. Of course, read speeds are where a RIAD 0 setup shines and if they hadn’t been better than what a regular SSD was capable of we would have been disappointed. It may not be as good as the X-25 but then again the X25 is set up in such away that you have massive parallel lanes which act a heck of a lot like a massive 10 drive RAID 0 setup.

 

[AkG]

Write Performance

Write Performance

For this benchmark HD Tune Pro was used. To run the write benchmark on a drive, you must first remove all partitions from that drive and then and only then will it allow you to run this test. Unlike some other benchmarking utilities the HD Tune Pro writes across the full area of the drive, thus it easily shows any weakness a drive may have.

While most OS drives spend most of their times reading and not writing, the write speed of the drive does have a big impact on the stutter issue and how fast the drive feels.

These numbers are decent, but not great. It is unfortunate that this generation of Samsung based MLC drives is not as fast at writes as the previous generation (e.g. the Warp v2). On the positive side, the majority of write speeds were all above 50MB/s with only 5 quick dips below this (with the worst being 16.8). These spots are prime candidates for stuttering and / or early death…but we have to wonder if it is the NAND chips or the controller which is at fault. It seems that along with all the positives associated with RAID 0 we are going to see the negatives as well.

In a proper setup the RAID controller is not only powerful enough to handle the overhead associated with RAID stripes but is backed up by onboard cache. This is why dedicated RAID boards have their own fast CPU and their own memory. Neither is the case with this JMicron controller (as it appears to be anything but fast and powerful) and its write performance is negatively effected by it. However, with all but a few spots being above 50MB/s we are impressed with the increase on the bottom end of its performance; heck we will take average write speeds of 80s over 90s any day….as long as the majority of low ends are all greater than 50MB/s and not 20MB/s. In the real world this generation of drives should “feel” a lot faster because of its improved bottom end; and more importantly it should have a lot less stutters because of this. To us this makes these write numbers one for the win column.

 

[AkG]

Crystal DiskMark / Random Performance

Crystal DiskMark

Crystal DiskMark is designed to quickly test the performance of your hard drives. Currently, the program allows to measure sequential and random read/write speeds; and allows you to set the number of tests iterations to run. We left the number of tests at 5. When all 5 tests for a given section were run Crystal DiskMark then averages out all 5 numbers to give a result for that section.

Read

​

Judging from the above numbers, this SSD loves larger chunks of data. It would appear that its controllers do add a certain amount of overhead to the scenario. This is not overly surprising and we will be taking a closer look at this when we get to the ATTO and IOMeter results. In the mean time, all three sets of numbers will eat any spindle based drive for lunch, and 14+ MB/s on 4K chunks is still impressive.


Write

​

As we saw with the read speed test, the small 4k test is handicapped by the RAID 0 controller and dual JM602 setup, but it still better than older Samsung based MLC and SLC drives and any improvement is a good thing.


Random Access Time

To obtain the absolute, most accurate Random access time, h2benchw was used for this benchmark. This benchmark tests how quickly different areas of the drive’s memory can be accessed. A low number means that the drive space can be accessed quickly while a high number means that more time is taken trying to access different parts of the drive. To run this program, one must use a DOS prompt and tell it what sections of the test to run. While one could use “h2benchw 1 -english -s -tt "harddisk test" -w test” for example and just run the seek tests, we took the more complete approach and ran the full gamut of tests and then extracted the necessary information from the text file. This is the command line argument we used “h2benchw 1 -a -! -tt "harddisk drivetest" -w drivetest”. This tells the program to write all results in English, save them in drivetest txt file, do write and read tests and do it all on drive 1 (or the second drive found, with 0 being the OS drive).

​

Our hunch about the RAID's pros and cons rearing their ugly heads does seem to have some merit. Now don’t get us wrong, 0.21MS is still freaking phenomenal and anyone who says that they can tell the difference between .18 and .21MS (or even .09 and .21) in normal day to day tasks is trying to sell you something. These numbers are very, very good and when compared against an older SLC drive they are impressive; heck, when compared against a super fast 10,000rpm drive whose main claim to fame is its low latency…well the results aren’t pretty for the vRaptor.

 

[AkG]

SIS Sandra / ATTO Results

SIS Sandra

This test was run with the removable storage benchmark in Sandra XII Standard. All of the scores are calculated in operations per second and have been averaged out from the scores of 4 test runs.

​

While we are not overly fond of this synthetic benchmark, due to near random like inconsistency, it does appear Sandra loves the Titan. Scores this big is certainly impressive and while they may have very little bearing on real world results they do make for some pretty decent bragging rights.

ATTO Disk Benchmark

The ATTO disk benchmark tests the drives read and write speeds using gradually larger size files. For these tests, the ATTO program was set to run from its smallest to largest value (.5KB to 8192KB) and the total length was set to 256MB. The test program then spits out an extrapolated performance figure in megabytes per second.

Read

​

We do find it interesting that G. Skill has actually under estimated (or at the very least downplayed) the power of this drive. This drive is supposedly only rated for a maximum speed of 200MB/s but as you can clearly see it exceeds this number by about the half way mark of the test.

On the down side, at the low end of the spectrum (or to be precise the first four smallest chunks from .5 to 4k) the Titan is about as fast as the Warp V2. To put it another way, the low end of the spectrum is where the negatives associated with underpowered RAID controllers and higher latencies come out to play. In this range the Titan is tied for dead last….but it quickly stages a comeback. By the time we get to 32KB chunks of data the Titan has gone from last to second place and heck as early as 8KB chunks is the second fastest SSD we have tested….and this is where it stays, right at the number 2 slot. It may not be as fast as the X25 but it’s no slacker either, and these are some very good numbers.

Write

​

As with the Read speeds, G. Skill has once again downplayed or underrated the performance of this SSD. It easily beats its rated top write speed of 160MB/s in the last quarter and stays there right to the end. It is refreshing when companies are humble about their numbers and it really does tell you a lot about the company behind them.