For this test, we are utilizing our fastest server to test the performance of the Icy Dock drive bay converter. We test the Toshiba PX02SMF080 with the converter, and directly attached to the LSI 9300-8e.
We have overlaid the results of our random 4K write testing to illustrate the performance with the converter and a direct cable connection to the HBA. Each dot represents performance during one second of the test, and we record data for five minutes for each Queue Depth. This method allows us to measure performance variability with a high level of granularity. We can see that the majority of results are overlapping, and the performance remains consistent both with and without the converter.
The lines represent the average speed during the measurement window. We can see that the SSD actually has a faster speed with the converter, but a word of caution. The converter does not make the SSD faster, but allows the SSD to perform very well within its normal range of performance. Small differences such as these fit within the +/- tolerance of 5% for a typical test, as flash solutions tend to have a bit of variability, even in steady state conditions.
Our 4K random read test again points out that there is no difference in performance with the converter. The majority of the black dots are covered, due to the device offering the same level of performance with the converter as when directly attached. This is truly spectacular performance from both the SSD and the converter.
Our write percentage testing illustrates the varying performance of each solution with mixed workloads. The 100% column to the right is a pure write workload of the 4K file size, and 0% represents a pure 4K read workload.
This more demanding test shows the Icy Dock converter allowing the 12Gb/s SSD to deliver full performance. The Icy Dock converter aces our performance tests, illustrating that it does not affect performance of the SSD.