Articles: Storage

Bookmark and Share

Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 ]

Testbed Configuration

For our today’s SSD test session we put together a system on an Intel H67 based mainboard. This chipset provides support for two SATA 6 Gbit/s ports, which we use to connect the tested SSDs.

Taking into account that OCZ Vector is shooting for the very top of the SSD hierarchy, we decided to compare this promising newcomer with the fastest SSDs we currently have on hand. As a result, the new OCZ Vector will be competing against products based on SandForce controller (Intel SSD 520 and Corsair Force GS), new Corsair SSDs on LAMD LM87800 controller (Neutron GTX and Neutron), our irreplaceable OCZ Vertex 4 on Indilinx Everest 2 controller, very impressive proprietary Samsung SSD 830, and Plextor SSDs on Marvell 9174 controller (M5S) and Marvell 9187 controller (M5 Pro). All above mentioned drives used synchronous flash memory.  In particular, Corsair Neutron, Intel SSD 520, OCZ Vertex 4 and Plextor M5S use 25 nm IMFT memory with ONFI-interface. And Corsair Force GS, Corsair Neutron GTX, Plextor M5 Pro and Samsung 830 use Toggle Mode MLC NAND manufactured using 2x nm or 19 nm technology.

Overall our testbed was configured as follows:

  • Intel Core i5-2400 (Sandy Bridge, 4 cores, 3.1 GHz, EIST and Turbo Boost turned off);
  • Foxconn H67S mainboard (BIOS A41F1P03);
  • 2 x 2 GB DDR3-1333 SDRAM DIMM 9-9-9-24-1T;
  • Crucial m4 256 GB system disk (CT256M4SSD2);
  • Tested SSDs:
    • Corsair Force GS Series 240 GB (CSSD-F240GBGS-BK, firmware version 5.03);
    • Corsair Neutron GTX 240 GB (CSSD-N240GBGTX-BK, firmware version 2.06);
    • Corsair Neutron 240 GB (CSSD-N240GB3-BK, firmware version 2.06);
    • Intel SSD 520 240 GB (SSDSC2CW240A3K5, firmware version 400i);
    • OCZ Vertex 4 256 GB (VTX4-25SAT3-256G, firmware version 1.5);
    • OCZ Vector 256 GB (VTR1-25SAT3-256G, firmware version 1.02);
    • Plextor M5S 256 GB (PX-256M5S, firmware version 1.00);
    • Plextor M5 Pro 256 GB (PX-256M5P, firmware version 1.01);
    • Samsung SSD 830 256 GB (MZ-7PC256D, firmware version CXM03B1Q).
  • Microsoft Windows 7 SP1 Ultimate x64
  • Drivers:
    • Intel Chipset Driver;
    • Intel Graphics Media Accelerator Driver;
    • Intel Rapid Storage Technology


Random and Sequential Read/Write

We use CrystalDiskMark 3.0.1 benchmark to test the random- and sequential read and write speed. This benchmark is convenient to work with as it can measure the speed of an SSD with both incompressible random and fully compressible recurring data. This feature is important for testing SSDs based on SF-2281/2282 controller, which tries to compress the data before writing it into the flash-memory. So, there are two numbers in the diagrams below that reflect the maximum and minimum SSD speed. The real-life performance of an SSD is going to be in-between those two numbers depending on how effective the controller data compression is.

Note that the performance tests in this section refer to SSDs in their “Fresh Out-of-Box” state (FOB). No degradation could have taken place yet.

It’s getting harder and harder to develop high-speed SSDs that would stand out among the competition. Moreover, modern SSDs have got so fast that they already feel constrained by today's SATA 6 Gbit/s platforms. It is the SATA bandwidth rather than their design that seems to limit the performance of modern SSDs at sequential operations. That's why SSD controller developers have shifted their focus to improving performance at random-address operations but there's another limiting factor here. The platform’s CPU may just be not fast enough to generate disk requests at the same rate as the SSD can execute them.

That said, OCZ has indeed come up with a highly promising solution. Compared to the other modern products, the Vector features a high speed of sequential writing and also offers good performance with a long queue of random-address requests. These are the parameters in which today's SSDs can hope to be different from their competitors. Therefore, the Vector’s results in the synthetic benchmark CrystalDiskMark 3.0.1 suggest that we’re dealing with a truly impressive SATA3 product.

There is only one nuance that you should know about. OCZ Vertex 4 drives were found to slow down when writing data continuously. When more than half their capacity gets filled with data, their performance drops dramatically. Fortunately, this is just a temporary problem which disappears in a few minutes. It is a just a side effect of the technology employed in OCZ drives to improve write speed in real-life applications.

This technology is employed in the Vector series, too. That’s why the graph of sequential write speed is the same as with the Vertex 4 if we use the whole capacity of the SSD:

So we see the Vector slow down after half its capacity if filled by continuous writing. The situation is rather unrealistic, though. Moreover, the speed returns to normal after a brief pause for the SSD to do some housekeeping. Any performance hits of this kind cannot be observed with the Vector in real-life applications.

This can be easily confirmed by a simple test with AS SSD version 1.6.4237.30508. Here is the speed of a clean 256 GB Vector:

After 128 GB of data is written to it (about 54% of the full capacity) in a single continuous session, we see the expected performance hit:

However, after a few minutes’ pause, the Vector is as fast as new.


Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 ]


Comments currently: 83
Discussion started: 11/27/12 03:03:47 PM
Latest comment: 11/12/16 11:20:59 AM

View comments

Add your Comment