Results: Random And Sequential Performance
Sequential Performance
Fantastic sequential read and write performance is a trademark of modern SSDs. To measure it, we use incompressible data over a 16 GB LBA space, and then test at queue depths from one to 16. We're reporting these numbers in binary (where 1 KB equals 1024 bytes) instead of decimal numbers (where 1 KB is 1000 bytes). When necessary, we also limit the scale of the chart to enhance readability.
128 KB Sequential Read
Plextor uses CrystalDiskMark to generate the 520 MB/s sequential read specification for its M6 drives. CDM uses neither 128 KB accesses nor reports in binary, so the 500 MB/s we see is more or less comparable. That's not as much throughput as the M5 Pro, however.
128 KB Sequential Write
The same applies to sequential writes, though the pair of M6es seems to do more with less. Using half as many dies, the new drives nip at the M5P's heels.
There are some gains from the controller and firmware, and changes to the flash. But that's enough to swamp the 256 GB Adata SP920 I just reviewed. Using the same number of dies and a similar controller, the SP920 only managed 344 MB/s. Despite similarities elsewhere, this illustrates Toshiba's A19 flash giving Micron's 128 Gb L85A NAND a run for its money.
Random Performance
We turn to Iometer as our synthetic metric of choice for testing 4 KB random performance. Technically, "random" translates to a consecutive access that occurs more than one sector away. On a mechanical hard disk, this can lead to significant latencies that hammer performance. Spinning media simply handles sequential accesses much better than random ones, since the heads don't have to be physically repositioned. With SSDs, the random/sequential access distinction is much less relevant. Data are put wherever the controller wants it, so the idea that the operating system sees one piece of information next to another is mostly just an illusion.
Random 4 KB Reads
The pair of M6es fall behind in the middle of our chart, which you especially see with 16 outstanding commands. Both SSDs recover to hit Plextor's rating of 94,000 IOPS by the end, though. Its M5 Pro is 3000 IOPS-faster up top thanks to the Xtreme firmware, but all three drives start from the same 8000 IOPS-range at a queue depth of one. And that's what matters. Random reads on the desktop typically happen at low queue depths, and 9000 to 10,000 IOPS is as much as you can expect from incumbents like Samsung's 840 EVO. Plextor's drives are in good company down there.
Random 4 KB Writes
Plextor says its 256 GB models should hit 80,000 IOPS in a 4 KB random write test, and that's exactly what we record in Iometer. Again, the older M5 Pro delivers 3000 additional IOPS at a queue depth of 32. All three SSDs perform very similarly otherwise, though.
