|OCZ RevoDrive X2 PCI-Express SSD|
|Reviews - Featured Reviews: Storage|
|Written by Olin Coles|
|Friday, 19 November 2010|
Page 10 of 12
AIDA64 Disk Benchmark
When development ended for Lavalys EVEREST the team picked up with AIDA64, an industry leading system diagnostics and benchmarking solution for enthusiasts PC users. Very few are aware of the Disk Benchmark tool available inside the program, which operates entirely on x64 instruction. The AIDA64 Disk Benchmark performs linear read and write bandwidth tests on each drive, and can be configured to use file chunk sizes up to 1MB (which speeds up testing and minimizes jitter in the waveform). Because of the full sector-by-sector nature of linear testing, Benchmark Reviews endorses this method for testing SSD products, as detailed in our Solid State Drive Benchmark Performance Testing article. However, Hard Disk Drive products suffer a lower average bandwidth as the capacity draws linear read/write speed down into the inner-portion of the disk platter. EVEREST Disk Benchmark does not require a partition to be present for testing, so all of our benchmarks are completed prior to drive formatting.
Linear disk benchmarks are superior bandwidth speed tools in my opinion, because they scan from the first physical sector to the last. A side affect of many linear write-performance test tools is that the data is erased as it writes to every sector on the drive. Normally this isn't an issue, but it has been shown that partition table alignment will occasionally play a role in overall SSD performance (HDDs don't suffer this problem). The high-performance storage products tested with AIDA64 Disk Benchmark use a 1MB block size option.
SPECIAL NOTE: Before we look over the results, it's important to mention that the RevoDrive series cannot make use of TRIM support because of the Silicon Image RAID controller. Standard measures to manually reset the NAND condition are not possible because of this architecture, and erase tools do not have the same effect. As a direct consequence of this, the SandForce SF1200 controllers are forced to manage write-block cleaning themselves. With the help of OCZ, we've discovered this could take weeks or longer depending on NAND density and write cycles. Due to the full sector-by-sector write process by AIDA64, the RevoDrive X2 demonstrated serious performance degradation near the end of the first test and in all subsequent tests.
Beginning with linear read performance, the OCZ RevoDrive X2 PCI-E SSD measured an average 755.7 MBps with a maximum peak speed of 774.4 MBps. This easily surpasses the original RevoDrive, which produced 510.5 MBps read speed and peaked at 511.6 MBps. Linear write-to tests were next...
The waveform chart below illustrates how well the integrated quad-SandForce SF-1200 DuraWrite buffers on the Revo SSD managed file transfers, and made linear write performance appears relatively even. This is even more impressive when you consider that this drive lacks TRIM support, and these tests followed after some of the others. Since the other tests use spot-testing methods to determine performance, the SandForce SSD controllers route our AIDA64 writes to 'clean' NAND sectors. This works well until it reaches the end of the run, where performance drops off.
This 240GB RevoDrive X2 produced an average linear write-to speed of 629, with a maximum performance of 665.1 MBps. The NAND condition in this test forces the X2 to compare nicely against the original RevoDrive, which recorded write-to speeds of 422.8 MBps with a 429.6 MBps maximum. For comparison, a set of SandForce-driven SSDs in RAID-0 produced 437.7 MBps read speeds, and 470.5 MBps with larger 128KB RAID-0 stripe sizes. The OCZ Revo PCI-Express SSD was much faster at read and write transactions.
The chart below shows the average linear read and write bandwidth speeds for a cross-section of storage devices tested with AIDA64 :
Linear tests are an important tool for comparing bandwidth speed between storage products - although HDD products suffer performance degradation over the span of their areal storage capacity. Linear bandwidth certainly benefits the Solid State Drive, since there's very little fluctuation in transfer speed. This is because Hard Disk Drive products decline in performance as the spindle reaches the inner-most sectors on the magnetic platter, away from the fast outer edge.
In the next section I share my final thoughts on the struggle between SSD and HDD technology, as well as their new hybrid counterpart, before delivering my conclusion and final product rating.