I finally got around to testing these drives in my chassis and so far have found similar results to those posted here.
For example, here are the results of a test with disk caching turned off:
-----------------------------------------------------------------------
CrystalDiskMark 5.2.2 x64 (C) 2007-2017 hiyohiyo
Crystal Dew World :
Crystal Dew World
-----------------------------------------------------------------------
* MB/s = 1,000,000 bytes/s [SATA/600 = 600,000,000 bytes/s]
* KB = 1000 bytes, KiB = 1024 bytes
Sequential Read (Q= 32,T= 1) : 161.983 MB/s
Sequential Write (Q= 32,T= 1) : 160.129 MB/s
Random Read 4KiB (Q= 32,T= 1) : 2.447 MB/s [ 597.4 IOPS]
Random Write 4KiB (Q= 32,T= 1) : 2.095 MB/s [ 511.5 IOPS]
Sequential Read (T= 1) : 162.090 MB/s
Sequential Write (T= 1) : 70.885 MB/s
Random Read 4KiB (Q= 1,T= 1) : 0.732 MB/s [ 178.7 IOPS]
Random Write 4KiB (Q= 1,T= 1) : 0.714 MB/s [ 174.3 IOPS]
Test : 1024 MiB [K: 0.0% (0.3/2794.4 GiB)] (x5) [Interval=5 sec]
Date : 2018/03/18 11:26:47
OS : Windows Server 2016 [10.0 Build 14393] (x64)
Whereas, here are the results of a test with disk caching turned on:
-----------------------------------------------------------------------
CrystalDiskMark 5.2.2 x64 (C) 2007-2017 hiyohiyo
Crystal Dew World :
Crystal Dew World
-----------------------------------------------------------------------
* MB/s = 1,000,000 bytes/s [SATA/600 = 600,000,000 bytes/s]
* KB = 1000 bytes, KiB = 1024 bytes
Sequential Read (Q= 32,T= 1) : 162.917 MB/s
Sequential Write (Q= 32,T= 1) : 161.515 MB/s
Random Read 4KiB (Q= 32,T= 1) : 2.450 MB/s [ 598.1 IOPS]
Random Write 4KiB (Q= 32,T= 1) : 2.426 MB/s [ 592.3 IOPS]
Sequential Read (T= 1) : 161.916 MB/s
Sequential Write (T= 1) : 161.061 MB/s
Random Read 4KiB (Q= 1,T= 1) : 0.725 MB/s [ 177.0 IOPS]
Random Write 4KiB (Q= 1,T= 1) : 2.392 MB/s [ 584.0 IOPS]
Test : 1024 MiB [K: 0.0% (0.3/2794.4 GiB)] (x5) [Interval=5 sec]
Date : 2018/03/18 11:03:09
OS : Windows Server 2016 [10.0 Build 14393] (x64)
The main difference is that the Sequential Write and Random 4K Write performance increases quite a bit once disk caching is switched on, as reported by czl.
Also, I use Hard Disk Sentinel to test all my drives and the Reinitialize test runs very slow (around 17 MB/s) whether caching is turned on or off. From poking around some posts on their forum, it look HDS disables caching for its tests which would explain this speed I'm seeing:
Yes, Hard Disk Sentinel disables the cache in all possible ways, exactly to prevent using the cache, as it is important to read/write the actual sectors, instead of the cache. For this, it uses direct communication method for the testing which disables caching of reads/writes.
I'm currently running the last of 4 tests (Reinitialize) and it's estimated that it will take 48 hours to complete! For comparison, my 8TB reds usually averaged around 24 hours on the same test. Again, I'm using these in a backup server and will have disk caching turned on for all of them so I don't foresee any issues with these when it comes to performance.