Anvil Storage Utilities
Version and / or Patch Used: RC6
So, what is Anvil Storage Utilities? It is a storage benchmark for SSD's, and HDD's, where you can check and monitor your performance. The Standard Storage Benchmark performs a series of tests; you can run a full test, or just the read or the write test, or you can run a single test (i.e. 4K QD16).
The software is used several different ways, and to show different aspects for each drive. We've chosen to use this software to show the performance of a drive with two different data sets. The first is with compressible data, and the second data set is incompressible data. Several users have requested this data in our SSD reviews.
0-Fill Compressible Data
This is actually the worst score I can remember for a 2 drive SATA III SSD array. However, upon further examination, we can see that most of what is bringing down the score total is coming from the bottom two write categories. So much of an Anvil's total score is predicated on these two categories, that when a drive/array does not pump out a lot of write IOPS at higher queue depths, the total score is drastically affected. We believe there are two factors in play here:
1) Toshiba has tuned the Q Series Pro for write performance at low QD
2) The lack of a DRAM buffer is drastically affecting synthetic based write performance.
0-Fill Compressible Read IOPS through Queue Depth Scale
Compressible 4K read performance at QD 1 - QD 2 is lower than what we typically see from a hyper-class SSD, but as queue depth stacks up, the Q Series Pro finishes strong.
0-Fill Compressible Write IOPS through Queue Depth Scale
The Q series Pro starts off lower than the rest of the arrays on our chart, and drops off of a cliff at 4K QD4. But remember: this is synthetic testing.
Compression 100% Incompressible Data
Incompressible scoring is pretty much the same as compressible scoring.
Incompressible Read IOPS through Queue Depth Scale
Read IOPS charting of incompressible data mirrors that of compressible.
Incompressible Write IOPS through Queue Scale
The variation we're seeing here is more a product of an inconsistent synthetic based performance curve due to the lack of a DRAM buffer, than a change in data types. Normally when we see this, it does not bode well for our test subject when we move to the testing that matters.