by Aleksey Meyev
07/01/2010 | 10:01 AM
In our recent roundup of solid state drives we noted the rapid development of that market and promised to test more such products. So today we are going to benchmark as many as ten SSDs from OCZ. This brand is well known to every computer enthusiast and its SSDs quickly earned a good reputation thanks to their efficient Indilinx Barefoot controller. We talked about them about a year ago, but there have been changes since then. OCZ has released new firmware and new product models while we have transitioned to a new testbed and methods. So, it is high time we took another look at the considerably changed line-up of OCZ’s solid state drives.
The mainstream Agility series is expected to offer the most appealing price/performance ratio. Every Agility model is based on a full-featured Indilinx Barefoot IDX110M00 controller which works with MLC flash memory and 64 megabytes of cache. OCZ declares identical speeds for both the 60GB and 120GB models: up to 230 MBps at reading and up to 135 MBps at writing (and a sustained write speed of 80 MBps). There is also a 250GB Agility with the same speed specs and a 30GB Agility with a read speed of 185 MBps and a write speed of 100 MBps. These SSDs were among the first to support the TRIM command.
Like the rest of the Indilinx-based SSDs in this review, these products have firmware version 1.5.
Most of today’s SSDs are designed in the 2.5-inch form-factor but this model is a heavy-looking 3.5-inch brick. We’ve got a 120GB drive whereas the top model in this series has a storage capacity of 1 terabyte! How did they achieve that? Well, there are in fact a few full-featured Indilinx-based SSDs inside the case and they all seem to be combined into a RAID0 array using an additional controller. Consequently, the total amount of cache available to the controllers is as large as 128 megabytes. The specified speed is up to 260 MBps both at writing and reading. However, the sustained write speed of our model is only 140 MBps compared to 220 MBps of its larger-capacity counterparts. This may be due to the senior models representing a 4-channel array (otherwise, it wouldn’t be possible to achieve a capacity of 1 terabyte with Indilinx controllers) whereas the junior model is a 2-channel array. This is but our supposition, though.
The Colossus Lite series has a twin which lacks the word Lite in its name and is positioned as an Enterprise rather than Performance one. The descriptions do not make the difference between them clear except for the speed of writing small data blocks and a slightly higher MTBF of the senior series.
This SSD does not support TRIM which must be the consequence of the additional RAID controller.
The firmware version is 1.0000.
The value Onyx series comes in capacities of 32 and 64 gigabytes and is positioned as affordable boot disks (by the way, we covered an opponent to this model, Intel X25-V, in our previous article). The Onyx is based on the well-known Indilinx Barefoot controller with 64 megabytes of cache. But as far as we know, the number of controller channels has been reduced due to the small amounts of flash memory. Together with some other possible changes, this has affected the speed specs. The Onyx has a peak read speed of 125 MBps (135 MBps for the 64GB model) and a peak write speed of 70 MBps. It supports TRIM.
As opposed to the other SSDs in this review, this model has firmware version 1.51.
The Solid 2 series is affordable, too. It is similar to the above-described Onyx but its top capacity is a little higher. It has an updated Indilinx controller with 64 MB of cache inside. The 30GB models from the Onyx and Solid 2 series have the same speed specs, but the 60GB Solid 2 is faster than its Onyx counterpart, promising a read speed up to 200 MBps and a write speed up to 120 MBps. The 120MB Solid 2 model is quite a serious one already as its specified read and write speeds are as high as 220 MBps and 130 MBps, respectively.
Interestingly, the specified capacity of the Onyx is 32 gigabytes (which is a power of 2 when measured in gigabytes) whereas the capacity of the Solid 2 series models is a multiple of 10 gigabytes. But when you take a look at the real formatted capacity of the junior models of these two series, you will find them to be the same at 30533 megabytes. Misleading the customer like that doesn’t look pretty.
The Vertex series is known for being the first SSDs from OCZ to feature the Indilinx controller. They are not going to leave the scene yet, so we will compare them with the newcomers. The 60GB model is declared to have a read speed up to 230 MBps and a write speed up to 80 MBps. The larger-capacity products are promised to deliver 250 MBps at reading and 180 MBps at writing. Interestingly, the 60 and 120GB models have somewhat lower write speeds than the 30 and 90GB ones, respectively.
Every new SSD controller is met with lots of rumors and discussions, especially if it is declared to have high speed specs. The SandForce controllers are just like that. Our Vertex 2 is based on the SF-1200. The other SandForce chip, SF-1500, is similar in design but is meant for corporate applications. Thus, it supports not only MLC but also SLC chips and, supposedly, has twice the number of memory access channels (16 rather than 8). Both controllers represent a fundamentally new design which lacks any cache memory (we mean external cache because the controller chip has an integrated cache but its capacity is not declared). The developer claims that the controller is far more tolerant to data loss than the opponents. Unfortunately, the details are not revealed although the controller seems to write data with redundancy encoding. Anyway, extra protection is always welcome. OCZ declares an MTBF of 2 million hours for the Vertex 2 series which is twice the MTBF of the other drives.
The downside of this controller is that it needs more flash memory for its housekeeping purposes. While most controllers need a reserve of 6.5% of the total capacity (this is why we have storage capacities of 30, 60, 120 gigabytes although the total capacity of the flash memory chips in them is 32, 64, 128 gigabytes), the basic line of storage capacities in this case goes like 50-100-200-400 gigabytes. It is easy to calculate the capacity loss, which equals 25%. After releasing such early models based on this controller, the manufacturers quickly offered 60-120-240-480GB products in which some of the capacity was returned to users at the (hopefully, reasonable) expense of reliability.
So, the 120GB OCZ Vertex 2 is specified to have a peak read speed of 285 MBps and a peak write speed of 275 MBps. Its sustained write speed is 250 MBps. Interestingly, the larger-capacity models are declared to have lower rather than higher speeds: 250, 240 and 200 MBps, respectively. Like every other modern SSD, this one supports TRIM (it has become a standard feature of all new controllers). OCZ puts a special emphasis on the fact that the firmware of this SSD is optimized for maximum performance in terms of operations per second.
This SSD seems to have firmware version 1.00.
The 30GB Vertex Turbo is the last product on our list. It differs from the Vertex series in the clock rate of its cache memory, which is 180 MHz. We could not see an effect of this overclocking before, but perhaps things are different now with new and optimized firmware? This model is specified to have a read speed of 240 MBps and a write speed of 145 MBps. The sustained write speed is 100 MBps. These are the specs of our model whereas its 120GB cousin is declared to deliver 270, 200 and 120 MBps, respectively.
The following testing utilities were used:
The SSDs were tested with the generic OS drivers and formatted in NTFS (wherever formatting was required) as one partition with the default cluster size. 32-gigabyte NTFS partitions with the default cluster size were created for FC-Test (if the drive is smaller than 64 gigabytes, it is partitioned in two halves). Each SSD was connected to a mainboard port and worked with enabled AHCI. We want to remind you that we have transitioned to a new testing methodology.
IOMeter is sending a stream of read and write requests with a request queue depth of 4. The size of the requested data block changes each minute, so that we could see the dependence of the drive’s sequential read/write speed on the size of the data block. This test is indicative of the maximum speed the drive can achieve.
The numeric data can be viewed in tables. We will be discussing graphs and diagrams.
SSDs have long been all right in terms of sequential speeds. Most of these models are limited by the interface bandwidth and far faster than hard disk drives. It is only the two 30GB products, the affordable Onyx and Solid 2, which cannot reach a speed of 200 MBps. The Solid 2 model is nearly 50 MBps faster than its cousin, though. The 30GB Vertex Turbo is true to its name and performs as fast as the 60GB products.
The Vertex 2 behaves in an interesting way. Its top speed is somewhat lower than 250 MBps and it accelerates to that speed somewhat later than the rest of the SSDs. Instead, it offers the highest performance on data blocks of 2 to 32 kilobytes.
The sequential write graphs produce a lot of information. First, we can note a peak in each graph which indicates the moment when the size of the written data block coincides with the size of internal blocks considering the number of controller channels. As a result, most of the Indilinx-based SSDs have a peak at 16KB blocks. The Colossus Lite prefers 64KB chunks of data, which may be a peculiarity of its firmware or due to its having four times the number of channels of a single controller.
Second, the Vertex 2 fails to meet our expectations based on its specs, even though its speed of 130 MBps is quite high in itself. Well, this SSD will have a chance to show a better performance in FC-Test. Right now, we can note that it is better than its opponents on very small data blocks.
Third, we can see a clear correlation between write speed and storage capacity. For example, the 30GB models are the slowest ones, the Vertex Turbo being the only 30GB drive to deliver 100 MBps on some data blocks whereas the Solid 2 is no faster than 31 MBps. The Onyx looks far more interesting. The 60GB models pass this test as a solid group, their graphs lying above 100 MBps at large data blocks. The 120GB Solid 2 and Agility are faster than 150 MBps.
For 10 minutes IOMeter is sending a stream of requests to read and write 512-byte data blocks with a request queue of 1. The total of requests processed by each SSD is much larger than its cache, so we get a sustained response time that doesn’t depend on the SSD’s buffer size.
The Indilinx-based models all have the same read response time. The latest firmware has improved this parameter to 0.1 milliseconds. The Vertex 2 differs by a mere 0.03 milliseconds, but this small difference may translate into a considerable amount of operations per second.
As for writing, it is easy to see that large-capacity SSDs have a better response time. The models of different series are not similar, though. The Vertex Turbo is as fast as the 60GB products. The Onyx is slightly slower and the 30GB Solid 2 is twice as slow as the 60GB Solid 2.
Now we will see how the performance of the drives in random read and write modes depends on the size of the requested data block.
The random read results roughly correspond to the read response time of the drives except for the Vertex 2 whose graph differs strikingly from the others. It is faster than the others on rather large data blocks but its performance does not scale up as rapidly as theirs on data blocks smaller than 8 KB and it falls behind its opponents as the consequence. Perhaps this is due to those mysterious ways of data processing which are meant to make the SSD more reliable.
Random writing repeats the results of the response time test, too. The only exception is the OCZ Vertex Turbo which is as fast as the 60GB models processing 512-byte requests. However, it falls behind the rest of the SSDs as soon as it has to process larger data blocks.
In the Database pattern the drive is processing a stream of requests to read and write 8KB random-address data blocks. The ratio of read to write requests is changing from 0% to 100% with a step of 10% throughout the test while the request queue depth varies from 1 to 256.
You can click the following links to view the tabled results for IOMeter: Database:
We will build diagrams for request queue depths of 1, 16 and 256.
The SSDs split up into several groups here. The Colossus Lite, 120GB Agility and Solid 2 are contending for top place, the latter two models producing nearly identical graphs.
Next goes the Vertex 2 which is inferior to the leaders but also much better at writing than the smaller-capacity models. Its weak spot cannot be hidden, though. It cannot match the Indilinx-based products when it comes to pure reading.
The 60GB drives pass this test close to each other, their graphs nearly coinciding, whereas the 30GB models differ. The Onyx and the Vertex Turbo are roughly equals while the Solid 2 is slower, especially at about 10% writes.
When there is a request queue, the SSDs enable their NCQ algorithms. As they support but rather short queues, we have very similar pictures at queue depths of 16 and 256 requests. The Vertex 2 goes ahead, proving the superiority of its controller. It effectively reorders requests in the queue in such a way as to load as many memory access channels as possible at the same time. The Colossus Lite doesn’t look good at high percentages of reads where its performance is rather low. It is just as good as the other 120GB models at high percentages of writes, though.
The SSDs split up into groups based on their capacity again, the Solid 2 still being the slowest model in the lowest-capacity group.
Winding up this part of our tests, we want to show you diagrams that illustrate each SSD’s performance at five different request queue depths.
This group of SSDs all behave in the same way as they share the same controller and firmware. The differences are due to the different capacities as we have already found out. The smaller-capacity products are worse at writing and their graphs go lower in the right part of the diagram.
The graph of the Colossus Lite is indicative of its using an array of controllers. It does not benefit as much as its opponents from the request queue. We don’t know if it is due to some firmware deficiencies (the earlier firmware versions, e.g. 1.31, used to behave like that) or to the combining of the controllers into an array, but this SSD is not very good in its current form. However, we’ve got the junior model of the series and it is quite possible that large-capacity Colossus Lite drives deliver outstanding performance unachievable with SSDs based on one controller only.
Based on the SandForce SF-1200 controller, the OCZ Vertex 2 differs dramatically from its predecessors and behaves not unlike Intel SSDs. It accelerates at long request queue depths, which is indicative of its ability to correctly work with a longer queue than the other SSDs support. Take note that the graphs for requests queue depths of 4 and 16 differ a lot, although they coincide with the Indilinx-based models. It is also good that not only the speed of reading but also of writing grows up at longer queue depths.
The drives are tested under loads typical of servers. The names of the patterns are self-explanatory. The results are presented as performance ratings which are calculated as the average speed of the drive at every load.
These graphs show us the longest request queue the drives support. It is the point where their performance stops to grow up linearly. Most of the SSDs have this point at a queue depth of 8 requests whereas the Vertex 2, at 16 requests. The latter’s graph has a bend that indicates a firmware flaw, yet this SSD is unrivalled, anyway.
The overall standings agree with the results of the previous tests. The drives that had a good read speed in IOMeter: Database cope with this test successfully, too. We can only note that the Onyx takes last place, being much slower than the other small-capacity products, and the Colossus has a low performance rating due to its small performance growth at long request queue depths.
The Vertex 2 enjoys an overwhelming advantage when there appear write requests. As opposed to the others, its performance grows up at longer request queue depths rather than remains the same, as that of most Indilinx-based models, or lowers as that of the Onyx and Vertex Turbo. The latter two are still faster than the Solid 2, though.
The multithreaded tests simulate a situation when there are one to four clients accessing the hard disk at the same time – the clients’ address zones do not overlap. We will discuss diagrams for a request queue of 1 as the most illustrative ones. When the queue is 2 or more requests long, the speed doesn’t depend much on the number of applications. You can also click the following links for the full results:
We’ve got interesting results at multithreaded reading. SSDs do not have read/write heads, so it shouldn’t really matter for them how many data threads they have to read. However, there is a difference as you can see. These SSDs are not so good at reading when they have to read multiple data threads. Nearly each of them suffers a 25% performance hit. The Vertex 2 comes out the winner in this test because its controller is more efficient at two and three data threads than the opponents.
We must also note one interesting fact. Every Indilinx-based SSD speeds up at four data threads. This controller must be able to effectively parallel such load to its memory access channels. This does not refer to the Colossus Lite. Its array of controllers limits the multithreaded read speed of this SSD to about 150 MBps.
Multithreaded writing is more difficult for SSDs (as opposed to HDDs that find it easier to do multithreaded writing than multithreaded reading). Almost all of our SSDs suffer a twofold performance hit when we add a second data thread for them to write, the Colossus Lite even slowing down fourfold! Interestingly, the Vertex 2 feels better in this test than its opponents even though it does not have 64 megabytes of cache that might smooth out the performance hit. It does not mind writing two data threads but has problems writing three threads. The fourth thread makes it even a little bit faster, though.
So, you shouldn’t write multiple data threads simultaneously to an SSD. It’s better to use the effective deferred writing mechanisms of modern OSes. But if you can’t avoid this kind of load, you should prefer the Vertex 2.
For this test two 32GB partitions are created on the drive and formatted in NTFS. A file-set is then created, read from the drive, copied within the same partition and copied into another partition. The time taken to perform these operations is measured and the speed of the drive is calculated. The Windows and Programs file-sets consist of a large number of small files whereas the other three patterns (ISO, MP3, and Install) include a few large files each.
You should be aware that the copying test not only indicates the speed of copying within the same disk but is also indicative of the latter’s behavior under complex load. In fact, the disk is processing two data threads then, one for reading and another for writing.
This test produces lots of results, so we will only discuss the Install, ISO and Programs patterns. Click the following link for the complete results table for FC-Test.
The Vertex 2 is slower than the other 120GB models, going ahead when processing small files only. The OS’s file caching mechanisms affect the results even though our testbed has only 1 gigabytes of system memory (if there are more memory, the OS is quite capable of putting the whole test pattern into it, reporting fantastic speeds of thousands of megabytes per second). As a result, the SSDs are faster with the Install pattern than in the ISO one which includes a large file that cannot be fitted into the OS cache.
The SSDs still split into multiple groups according to their capacity. There is only one exception: the Vertex Turbo is as fast as the 60GB group.
Every SSD is good at reading, especially when it comes to small files (you can compare this to the results we have in our HDD reviews). Still, some SSDs are better than others. For example, the Vertex 2 is slower than the 120GB models, just like in IOMeter’s tests of sequential operations. The Onyx is somewhat slower than the other 30GB SSDs.
The 120GB models win the test of copying. Their advantage is especially clear with large files. Interestingly, the Vertex Turbo is again ahead of the 30GB products and joins the larger-capacity ones.
Compared with the previous versions, the Vantage version of PCMark is more up-to-date and advanced in its selection of subtests as well as Windows Vista orientation. Each subtest runs ten times and the results of the ten runs are averaged.
Here is a brief description of each subtest:
Basing on these subtests, the drive’s overall performance rating is calculated.
We don’t have to discuss each test one by one because the overall performance ratings are illustrative enough. The 120GB models are barely ahead of the 60GB ones. The difference is really negligible. It is the advantage of the Vertex 2 over the others that provokes no doubt. Its performance in the individual tests and its overall performance rating is 50% higher than that of its opponents. SandForce can be proud of such an excellent controller that can score 41 thousand points in PCMark.
As for the 30GB models, the Vertex Turbo is true to its name and almost as fast as the larger-capacity products. The Onyx gives way to the Solid 2 and takes last place. We must note, however, that its overall score of 21.5 thousand points is higher than many SSDs from other brands can offer, the Solid 2 being an affordable model in which everything has been sacrificed for the sake of low price.
Next goes our homemade test of defragmentation speed. We created a very defragmented file system on a 32GB partition of a disk by loading it with music, video, games and applications. Then we saved a per-sector copy of the disk and now copy it to the disk we want to test. Next we run a script that evokes the integrated defragmenter of Windows 7 and marks the time of the beginning and end of the defragmentation process. For more information about this test, you can refer to this article.
We must remind you that defragmentation is useless and even harmless for solid state drives due to their operation principles. However, we use this test as it allows to benchmark their performance at a rather peculiar load that involves both reading and writing of small data blocks.
Our test image is exactly 32 gigabytes large, so we cannot benchmark the three 30GB models in this and next tests.
The SSDs pass this test with close results and are ranked up according to their capacity with only one exception: the Vertex 2 leaves the others behind and wins the test.
Now we are going to show you one more interesting test in which we use WinRAR version 3.91 to compress and then uncompress a 1.13GB folder with 8118 files in 671 subfolders. The files are documents and images in various formats. These operations are done on the tested drive. This test depends heavily on CPU performance, but the storage device affects its speed, too.
The results are close in this test, too. The SSDs do not even split into groups according to their capacity. However, the Vertex 2 manages to win the test of unzipping the archive. Considering the minor role of the storage device in this application, the advantage of 6 seconds looks impressive.
You can refer to our article called Hard Disk Drive Power Consumption Measurements: X-bit’s Methodology in Depth for details on this test. We will just list the specific modes we measure the power consumption in:
Let’s check out each mode one by one.
The results suggest that the higher the storage capacity of an SSD, the higher startup current it requires. A few interesting facts can be observed. First, some models, e.g. the Vertex Turbo, need a higher current than their cousins. The 60GB Agility needs even more power than the 120GB Agility! It is the Colossus Lite that needs the highest current to start up due to its internal design. The Vertex 2 has low power requirements.
The Colossus Lite needs over 1.5 watts when idle, which is three times as much as the other SSDs need. That’s a terrible appetite.
The Vertex 2 needs more power than the others, too, but its requirements are reasonable at 0.65 watts. The Onyx is the most economical SSD in this test.
When it comes to random-address operations, the Onyx is again the most economical and the Colossus Lite is the hungriest SSD. The 60GB Agility needs a lot of power at writing and the Vertex 2, at reading.
The Vertex 2 is the most economical drive at sequential writing, but is far from brilliant at sequential reading where it is only better than the Colossus Lite which has indecent power requirements. The 60GB Agility consumes an unexpectedly and inexplicably large amount of power when writing.
We will begin our Conclusion with the most exciting product in this review, the OCZ Vertex 2. Being the first SSD with a SandForce SD-1200 controller to be tested in our labs, it has proved to deliver excellent performance at loads typical of home computers. It is indeed very, very good. It is universal, too. In fact, we have not managed to find any flaws in it except that we did not see it deliver the specified read and write speeds in any of our tests. But we are ready to put up with that to have its high sequential speeds and very good performance in real-life applications. The lack of noticeable flaws in its firmware should also be mentioned.
As a result, we are proud to award OCZ Vertex 2 SSD with our Editor’s Choice title:
The low-capacity models are going to be interesting to many users who cannot afford the large ones. The Onyx, Solid 2 and Vertex Turbo are not so bad, after all. Yes, their sequential speeds are not very high, especially at writing, yet also not as low as to present any problems. Instead, these SSDs offer the main advantage of their class of devices: their extremely low response time at reading makes them excellent at any load with lots of requests with random or pseudo-random addresses. The Vertex Turbo is somewhat faster than the others, but also belongs to a more expensive series. The Onyx and the 30GB Solid 2 are roughly equal to each other, but the former is somewhat better at server loads while the latter, at operations with files and in home computer applications.
But if you can afford a large-capacity SSD, take it without thinking twice! Our today’s tests show that large-capacity SSDs are best at writing and mixed loads. They also have higher sequential speeds and better performance at server loads than their smaller-capacity counterparts.
We guess the only disappointment in this test session is the Colossus Lite. It didn’t show anything extraordinary in terms of performance, being often slower than same-capacity opponents, but proved to have high power consumption. However, it is possible that this SSD can only show its best in its maximum-capacity version: Colossus Lite is the only SSD series from OCZ to include a 1-terabyte model!