OCZ Vertex 3 240GB SSD Review

I/O Performance

There is little point of having an SSD drive that has blazing sustained reading and writing speeds, if the drive can’t handle reading and writing of small random files. If you intend to use your new SSD drive to store and run your operating system, then the drive must be able to cope with the many small random files that Windows will write to the drive continually. So I feel it is very important to test how many of these random files that a drive can handle in one second. I believe that anything over 1,000 I/O’s per second would be enough for most users running a consumer grade mainstream PC, and should provide a smooth running system. But obviously, the more I/O’s that a drive can handle, the faster the drive will feel and leave more headroom for those huge multitasking sessions that users sometimes engage in.

The things that I will look at are the total I/O per second and total MB/s.

Partition alignment and sector boundaries

Windows 7 will automatically align a partition to 4k boundaries during partition creation, Windows XP won’t. It is imperative that an SSD’s partition is aligned. Windows XP is also restricted to sector boundaries, while Windows 7 will use 4k boundaries if it can. The Vertex 3 is 4k boundary aware, and will use these boundaries when it can. Of course it will also remap LBAs for compatibility with the sector boundaries so the drive can be used with Windows XP.

IOMeter allows us to set the sector boundaries for conducting the tests, and I have therefore set the sector boundaries at 4K, which means the IOMeter tests are valid for Windows 7 and Windows Vista users. XP users will not be able to obtain such results.

I will provide a screenshot of the tests on the review drive for those of you who like to see the actual test result. All the comparison drive results are represented in the form of graphs.

If any of you would like to see a screenshot from any IOMeter test on a particular drive, please feel free to request one, and I’ll post the screenshot in the forum thread.

All the IOMeter tests create a 10GB data set on the target drive, and each test is run for a duration of 3 minutes.

IOMeter 4K random (outstanding I/Os = 4, 32) write test.

Our first test involves creating continual 4KB random files on the target drive with IOMeter. I use a 4KB file size, as it is believed that Windows will create and modify many of this size of file constantly in the background during a typical Windows session. I will use queue depths of 4 and 32 for these tests on the OCZ Vertex 3, OCZ Vertex 2, Plextor PX-256M2S, and Crucial RealSSD C300.

SATA 3

(Queue depth 4 SATA 3)

(Queue depth 32 SATA 3)

SATA 2

(Queue depth 4 SATA 2)

(Queue depth 4 SATA 2)

IOPS

MB/s

78878 4K random write IOPS is just outrageous performance from a single SATA SSD, and probably more than any desktop user will ever require. Nonetheless, it’s nice to have this much performance in reserve. The key to high 4K random performance is low latency, so even if Vertex 3 will never require queue depths of 32, its ultra low latency will also insure it’s faster than the other SSDs in these tests, regardless of how low the queue depth is.

4K random write queue depth profile

For this test I used various queue depths from 1 – 32 to give you an idea how this SSD performs at different queue depths. For a normal desktop user for lightweight multitasking, the queue depth will rarely rise above 2. For heavy multitasking, the queue depth is unlikely to rise above a value of 8.

Only two SSDs were tested here. The OCZ Vertex 3, and the Crucial RealSSD C300 128GB which was only tested on SATA 6Gbps, while the Vertex 3 was tested on SATA 3Gbps as well as SATA 6Gbps.

For this test I have also used completely random data, pseudo random data, and the normal repetitive data IOMeter test patterns. This has been done to show how the SandForce SF-2281 SSD processor deals with data that isn’t so easily compressed.

The result is below.

What is very interesting here is the “queue depth 2” result. With a queue depth 1 score of around 68MB/s even with perfect scaling, you wouldn’t expect the queue depth 2 result to be more than twice that of queue depth 1, and here we have a huge jump in performance at queue depth 2. You would think that some sort of glitch happened while running the queue depth 2 tests. This is what I first thought had happened, but the result is repeatable time after time, and the result is no more than 2MB/s out with each run of the test.

Could IOMeter itself be to blame?
If so, then why doesn’t the Vertex 3 get a similar boost on SATA 2, and why doesn’t the Crucial C300 get a similar boost on SATA 3?
Could AsRock or Intel have done something to tweak SATA 3 at queue depth 2?
Again, then why doesn’t the Crucial C300 get a similar hike in performance at queue depth 2?

The only plausible explanation that I can come up with is, SandForce have done some clever tweaking for 4K random write at a queue depth of 2, knowing that this area is where some real world benefits can be achieved.

Below I present a table of the results in more detail.

IOMeter 4K random (outstanding I/Os = 4, 32) Read test.

If there are many 4k files created, then that must also mean that many 4k files need to be read. This test measures 4k reading performance.

SATA 3

(Queue depth 4 SATA 3)

(Queue depth 32 SATA 3)

SATA 2

(Queue depth 4 SATA 2)

(Queue depth 32 SATA 2)

IOPS

MB/s

The Crucial RealSSD C300 was always mighty with 4K random reading performance, and is two times faster than the SF-1200 based Vertex 2. While the Vertex 3 still can’t match the performance of the RealSSD C300, it is getting closer.

4K random read queue depth profile.

This test is much simpler than the 4K random write test, as there is no need to include full random or pseudo random data.

Below I present a table of the results in more detail.

IOMeter 512KB (outstanding I/Os = 2) write test.

Sequential writing performance is also very important; in this test sequential writing performance is measured.

SATA 3

SATA 2

IOPS

MB/s

The OCZ Vertex 3 slaughters the competition in this test. And remember, the Vertex 3 is connected to a single SATA port here, so this isn’t relying on PCIe or multiple SSDs in a RAID 0 array, this a single SATA drive.

IOMeter 512KB (outstanding I/Os = 2) read test.

This test measures 512k sequential reading performance.

SATA 3

SATA 2

IOPS

MB/s

Sequential reading performance is very impressive, and the Vertex 3 is the fastest reading drive.

IOMeter Workstation simulation (outstanding I/Os = 64).

When running applications you will find that there is a mixture of small random files, and larger sequential files, being created and read. Not only that, it isn’t just one file at a time. In this test I measure a simulated workstation pattern, with a queue depth of 64 (threaded).

SATA 3

SATA 2

IOPS

MB/s

The OCZ Vertex 3 has done extremely well in the workstation simulation, and is a good deal faster than any of the competition.

Summary

The OCZ Vertex 3 and its SandForce SF-2281 SSD processor is really pretty much out on its own. In writing performance, be it small file random, or large sequential performance, it’s simply miles ahead of the competition. Reading performance is every bit as impressive, while it can’t quite manage to muster the might of the RealSSD C300 for small file random performance, it’s miles ahead in sequential reading performance.

Now let’s head to the next page for some real world tests….

 

• Synthetic benchmarks
• Real world tests