Review

Review: OCZ Apex Series 120GB SSD

The OCZ Apex Series of Solid State Drives (referred to as SSDs from now on) has been out for a few months now, and thanks to the kind people at OCZ, I've been able to get my hands on one of the 120GB models from the series. I will be reviewing both the product, and the technology, comparing the benchmark and test results from the SSD with the results from a traditional hard drive, but we will go into that later.

First of all, let's take a look at the SSD itself. This 120GB 2.5" SATA II SSD is available in the UK for just over £330 in most online stores, and uses MLC (multi-level cell) technology, as opposed to the more expensive SLC (single-level cell) technology. Naturally, the size of it makes it more appropriate for use in laptops and smaller computing devices; however with some easily available adapters it can be used in a desktop setup too.



The SSD is both small and light, as you would expect from a 2.5" drive, and, although I was unable to test it, the packaging claims that the device is shock resistant too.

The test system I will be using for this review is as follows:
Intel Q6600 Quad Core (overclocked to 3.0GHz)
OCZ PC2-8500 2GB DDR2 RAM
GIGABYTE GA-P35C-DS3R Motherboard
NVIDIA GeForce 8800GT 512MB
Thermaltake Kandalf – Water cooled (which also provides cooling for the CPU)

The drives we will be using are:
1x OCZ Apex Series 120GB SATA II Solid State Drive
1x Seagate Barracuda 256GB SATA II Hard Disk Drive

Each test was performed with only the drive being tested plugged in.

So without further a due, let's get started on our first test – the time it takes to install Windows Vista Home Premium. We will have technical benchmarks a little later, but this was the first thing that had to be done. The hard drives were formatted, and had new partitions (which took up the full capacity of the disk being tested) created on which Vista would be installed. The timer was started from the moment that the 'Install' button was pressed, to the first restart. The horizontal axis shows the time taken (in minutes) to complete this process.

So it's the first test and the OCZ SSD has lost out on this one, but only by a few minutes. At first I thought this was an anomalous result, but after repeated tests on both drives, this was the final result. This test was mainly a "just out of interest" one that could be described as weak anyway, however the SSD loses out in this one regardless.

After installing drivers and the tools I would use to benchmark the two drives, as well as all updates to Windows Vista (including Service Pack 1), the next test would be the PCMark Vantage HDD Suite benchmark. The results from the benchmark are gathered by performing numerous tests, both benchmarks and real-life scenarios, such as loading up music libraries and streaming media, whilst saving it at the same time. Now if the previous test gave the SSD a poor introduction, then this more than compensates for it. Below, you can clearly see that the OCZ Apex SSD races ahead.

As you can see, the SSD scores over 3 times higher in this particular benchmark than the traditional HDD, and this score is the result of numerous tests performed using typical applications too. However, next up is the HD Tune benchmark, which will put the numbers given on the back of the SSD to the test.

The OCZ Apex SSD claims to reach performance levels of up to 230MB/s for reading, and up to 160MB/s for writing. These speeds were, as the box puts it "Based on ATTO", however, even with that said, the speeds mentioned have to be taken with a pinch of salt, because quiet often the speeds are only achievable in ideal conditions.


The burst rate test works out the speed it takes to transfer data from the hard drive to the operating system. Interestingly the SSD lost out in this test. The next test measures the maximum, minimum and average reading speeds on the drive, and the SSD wins this one. The average read speed of the SSD is over 50% faster than the speed of the HDD.

The temperature of the drives during each benchmark was consistent. Surprisingly, the SSD remained at a steady 44 degrees Celsius, whereas the HDD remained at just 32 degrees Celsius. And this was the case in every benchmark, when the room was cool, and when the room was a little warmer. I noticed that the SSD was both quick to warm up (after boot up) and quick to cool down (after shut down), whereas the HDD was slower in both cases, but ultimately cooler over all.

Now, onto seek times. The seek time is what makes accessing data quick (or slow, as it may be), and so naturally, with no moving parts, SSDs have the upper hand when it comes to this. The following graph shows the seek times of the two drives.

The low seek times are what makes SSDs stand out when it comes to boot times and application loading times, however this is also requires the drive to have fast reading and writing speeds. In regards to writing speeds, the OCZ Apex SSD fell behind the HDD, reaching just under 100MB/s, losing to the HDD which gained just under 140MB/s. However, nearly all SSDs at the time of writing reach substantially lower write speeds than they do read speeds, so it is more of a problem with the technology than a problem with the product.

The final test shows the time taken to both start the computer and shut down the computer. The timer was started from the moment the power button was pressed, to the moment the login screen was displayed. In the case of the shut down test, it was from the moment the shut down button was clicked, until the power light went out. The time it took to go from the POST screen to the operating system loading screen was 20 seconds in each case, so I have provided the results with that time removed just for interest. Needless to say, in each chart, lower is better.

This was perhaps the most obvious advantage to having an SSD for me, personally. The boot time of the system using the SSD appears only a little different at first, but once the time taken to go from the POST screen to the loading screen is subtracted, the boot time for the SSD is pretty close to being half that of the HDD system. Shutdown times remain little different.

So far I have talked a lot about the technology behind this product as opposed to the product itself. The majority of the benchmarks show that the OCZ Apex SSD is substantially faster than the HDD, as well as showing some of the downsides to the technology. However, benchmarks don't count for everything.

Initially, before I installed the drive, I wondered if the drive would suffer from any of the stuttering issues that plagued the initial SSDs. Not once did I notice this. This was what put many off buying SSDs for a while, however, that brings me on to the main issue that put most people off: the price.

The model reviewed here, as mentioned before, costs around £330 in most online stores in the UK. The OCZ Apex 120GB SSD compares favourably (in terms of value) with the Intel X25-M 80GB, which, for roughly the same price offers only a faster read performance of 250MB/s, but a measly write speed of 70MB/s. Although the Intel range uses SLC technology, which means fewer errors in data transfers, it costs an unbelievable amount more. At the same time, the G.Skill Titan 128GB model is priced £50 less than the OCZ Apex we looked at today, yet offers all the same as it, except for the read performance which is 30MB/s less.


In my personal opinion, the OCZ Apex Series 120GB model offers a good compromise between value for money, and great performance. The mean time before failure is 1.5M hours, which is the same as many other SSDs in the market.

Although this SSD might offer better value for money than some SSDs in the market, £330 is still a lot of money to spend on a 120GB storage device, solid state or not. This is the unfortunate downside that comes with the technology, which limits larger capacity, higher performance SSDs like this to a select few, mainly enthusiasts.

The bottom line is, the OCZ Apex Series 120GB model is an expensive addition to your computer – there's no two ways about it. However, if you've got system, and you've got the money, you will love this excellent alternative to a traditional HDD, whether it is in your desktop or in your laptop.

Report a problem with article
Previous Story

Microsoft unveils blue 360 development kit

Next Story

Study: IE8's SmartScreen leads in malware protection

29 Comments

View more comments

yes I agree regarding the optimisations and the stuttering, however I may be wrong but I think they can suffer because they have no cache?

Although I know they are based on the same controller but now put in raid 0 the controller is far from the best, but I would still expect better performances than that.

Perhaps you should post this review over on their forum, those guys really know their stuff I'm sure they would be able to find out why you are not getting the full performance. I wouldn't expect the 240 MB/s but running at half that speed does seem a little slow.

It does suggest it is something to do with Vista because the bootup times are faster on those drives, so that suggest it is after bootup.

I wish I had the answer for you, I am running 2x Samsung 32gb SLC in Raid 0 and they are benching at just under 200 MB/s that is on the bad Raid built into the motherboard is well.

Yes, according to an article here:

http://www.pcper.com/article.php?aid=661

It uses a pair of them though (effectively creating a RAID partition). I could not see this for myself though, because taking apart the drive would mean voiding the warranty, and this is not my drive to take apart.

According to the article, it is a new version of them anyway.

nice reivew, i think you need to upgrade your first chart, the horizontal is seconds, i thought it was minutes?

Do these drives have ridiculous stutter like the previous cores and so and so?

Have you tried using it as your main drive and mulittasking, music playing, downloading and web browsing at the same time?

I've seen no stuttering issues as of yet - I've been using it for over 3 days now, but noticed none of those problems.

This is mainly due to the drive using two new and improved JMicron controllers in a RAID configuration (although you don't know it), which have resolved this issue.

I may be missing something here, but, in the article you state that the mean time before failure is 1.5M hours. I can't work out how they can have tested this, 1.5M hours (1,500,000 hours) is over 170 years! How can someone really say that their product has an average lifespan of over 170 years when the people who developed this technology weren't even born yet?

If anyone here can explain how a company can claim this, or explain a fairly logical way of being able to test this, I would love to hear it.

Probably just 'hardcore' testing it. Finding how much degradation occurs over that time, and multiply it to see how long it would take for the drive to fail.

Hi,

I'm going to be using the SSD as my main drive over the next few days, so I will be able to tell you then. Unfortunately some exams tommorow prevent me from doing it now. Are there any specific things you would like me to look at?

Sorry Neowin, but not testing the write speed of SSD is disingenuous at best ... write speed IS the major defect of SSD. Plus next time compared it to a good drive like a velociraptor...

The Apex is a real bad SSD, and to OCZ credit the new Vertex is MUCH better for all reviews I have seen so far.

Neowin get a F for that review.

Unfortunately, all the equipment used here is provided by the writers themselves. This is why I could not provide a better drive such as a Velociraptor, because the Seagate Barracuda is the best I have, and I simply cannot afford to buy a new, unneeded hard drive for a review, especially after buying a copy of PCMark Vantage.

Likewise, this review was completed in my own spare time, outside of school time. In order to measure the write speed of a drive, you have to remove all partitions (at least, using the tools available to me you did), and I have yet to have chance to test that.

I hope you can understand I have provided approximate write speeds in the review, because I cannot gain an accurate one unless I remove all partitions (as HD Tune requires).

However, I will try and get you some results as soon as possible.

EDIT: I did address this in the review, actually, not in the form of a graph, however. The maximum read speed using the writing benchmark is 150, however, as stated, it typically remains around 100MB/s.

Not meant as a personal attack at all... did not know the circumstances.

Should have stated that no better drive was available, and no write test could be done ... the Apex have pretty bad review in general, so it ticks me a bit I admit.

Sazz, if you could give it a good hammering with regards to mulitasking in real world use..

I.e. Browsing websites with say 4 tabs, listening to music in Windows media player, IM'ing on MSN/Digsby simultaneously...

And if you have oodles of time available - try all the above with http://www.eveonline.com running would be nice to know if it can cope or whether it will die in a fit of stuttering

Ha, funnily enough I bought a subscription to EVE last week, so I'll be happy to try it I doubt I'll find a way of benchmarking that, but I can certainly give it my best judgement.

I'll PM you the results when I get them, because it's likely to take me few days to get enough time together (especially as I have to re-install Vista, and all the hell that accompnies that process).

I'm very surprised about the results, seems to do pretty badly in everything.. I've seen some videos online and it shows the drive to be really fast.. *shrugs*

Not sure why OCZ sent you one if their crappy SSD's so late in the game. The Vertex is widely available and is the only real alternative to the X25M which is still regarded by many as the best SSD available today. If OCZ sent me an Apex for review today, I would be insulted.

And remember, benchmarks hardly tell the whole story. You aren't going to notice the difference in max read/write performance between drives, but you will notice the random read/write performance between drives, especially when using them as OS drives. This is where the Intel X25M screams and why it still feels faster in every day use as compared to the Vertex, even though the Vertex has bigger benchmark numbers. I own 5 different SSD's, none of which have a jmicron controller, so I speak from experience.

X25M or Vertex or go home (when considering MLC drives)

This was the first time I have spoken to OCZ, so it took a bit longer because I had to explain what we do, why we should have a review unit, and our benchmark results etc.

Now that they've had a previous dealing with us, it should take less time to get a review unit next time, fingers crossed ;)

And yes, I agree with what you say about the benchmarks. This review was meant to be just one to start things off though, as I have no other SSD to compare it to. Now that we've got one of these, I'll have something to compare what will hopefully be an Intel SSD (model unknown) next time.

Commenting is disabled on this article.