I think it's an Windows NTFS
My understanding is if you use Windows' NTFS, you're pretty much locked-in to a 4K boot partition cluster size. Some claim this is because Windows NTFS boot-up is hard-coded to a 4K size. From what I've seen that seems to be the case. Others say that some Windows background operations that access the boot partition will quickly corrupt if the cluster size is anything but 4K (e.g. some of the internal databases, etc).
I've read of attempts to "trick" the Windows NTFS OS into a different cluster size on the boot partition, but they seem dodgy to me. Examples:
Install on the default 4K cluster-size boot partition then backup and restore it onto a 32K cluster-size partition (I have not seen this work).
Create some kind of separate "sub-boot" partition where only a VERY small amount of "core" boot data are stored (supposedly about 50M
B). Somehow the OS installation is divided between this "core" partition and a second standard definable partition. Maybe
they were talking about using a specific boot manager? Memory is fuzzy on this one.
Replace a handful of OS files with older Windows OS files (yeah...that sounds smart
Like I said...dodgy. I quit messing with trying to change the boot
partition cluster-size years ago. I'll try to do some up-to-date research when I'm feeling better.
That's one reason why I minimize the size of my Windows boot partition on mechanical HDDs.
If I'm stuck with 4K clusters for the Windows NTFS OS itself, then to maximize performance in other areas, I want the system partition to be large enough to contain only the OS and its non-relocatable support files and allow for some expansion. For example, I make the boot partition (C disk) no larger than 40G
B (can even be much less) and store everything else on different physical drives
. On a current 1TB drive that 40GB partition is less than 5% of the drive, so to not waste a lot of the drive, use the remainder for low-access archive
files only. I also often move many (but not all) OS paging/data/log/email/etc files to a different physical drive
...although you have to be more flexible in your backup protocols when you do that. Many applications also by default store their data/log/temp files on the "C disk" (e.g. Office, browsers, etc) and some of them have a LOT of constant background activity on the drive...which affects OS performance. Those too often can be changed to different drives. My take is...in GENERAL
...the less "crap" you put on the "C disk", and the less "other" activity you place on the physical drive containing the "C disk" the better the performance of the OS.
realized I meant to distinguish between physical drives and partitions
sorry about the numerous edits. I'm giving up for the day.