When I was purchasing SSDs for my Home Lab, there wasn’t a lot of useful benchmarks out there for my use-case. I started out with the cheapest, they are all the same right? Wrong.
My main use-case is nested VMware Cloud Foundation which is already very resource intensive, but hammers the storage due to the write amplification of running nested vSAN.
Some write heavy operations like deploying OVAs were taking a very long time, so I started looking at performance on my Synology NAS.
Disk 6 – Samsung EVO 870 2TB (Faster)
Disk 7 – Samsung QVO 870 2TB (Slow)
Disk 8 – Samsung QVO 870 2TB (Slow)
You can see here, during periods of heavy writes, the QVOs are maxing out at ~40MB/sec while the EVO is sitting at a comfortable 25%. I pulled the QVOs and replaced with EVO.
Disk 6 – Samsung EVO 870 2TB
Disk 7 – Samsung EVO 870 2TB
Disk 8 – Samsung EVO 870 2TB
Now with all EVO, we’ve got a significant performance boost! ~120MB/sec compared with ~40MB/sec. Drive 7&8 are showing much lower utilization than 6 because they were brand new drives (SSDs operate a little slower once they’ve had some rewrite cycles).
I’m not going to go into the how/why, plenty of people have already written about that (see TLC vs QLC).
So choose your SSDs wisely, the cheapest ones will absolutely impact performance.