Can anyone explain why my 1TB SN850 is 930 GB and not 931 GB?

Associate
Joined
10 Dec 2007
Posts
1,784
Edit: Figured it out. Apparently this is because Windows has decent sized recovery partitions (500 MB+) putting it just below 931 GB. 931 GB is only for pure storage drives.

---------------------

I know this sounds funny to be questioning a 1GB difference but I'm just interested in the technical reasons or if there is something else going on.

Everywhere I read a 1TB drive should read as 931 GB in Windows due to the whole 1024 vs 1000 stuff. That's common knowledge.

However on my fresh Windows 10 PC build I have run into an issue with my SN850 1TB nvme. It's showing as 930GB under "This PC".

This guy says the same thing for his SN850 (last post): https://www.tenforums.com/general-s...-installations-c-drive-one-10gb-larger-2.html

Can anyone tell me why this specific nvme is not showing 931 GB like every other hard drive should? Can anyone else confirm?
 
Last edited:
Your drive should be around 837GB once you factor in over provisioning (the recommended standard is 7-10%). This keeps a chunk of the storage cells for wear maintenance etc. Any SSD of a sizeable amount, like 1TB+ and especially a heavily used drive like having an OS on it should be over-provisioned. with said 10%

It's as simple has having 10% of the drive left as unallocated space. The SSD controller will do the rest as and when.

Here's my 1TB NVMe for example:

AYKM3rN.jpg

Some useful reading on OP: https://www.minitool.com/partition-disk/ssd-over-provisioning.html

In short, set a minimum of 7% space that is not allocated using any disk tool you have, typically it's an option in the drive maker's tool like Samsung Magician, or just shave off 7% in Windows Disk Management. That's it.

Edit*
There is an argument to be made whether in a consumer use environment if OP is even necessary. The idea of it is as the storage space gets used, the drive will get slower as you reach the limit. OP alleviates that as the SSD controller can shuffle blocks from the OP area to offload maintenance and wear levelling etc so the user available space is always at a constant performance/reliability level.

The jury is out on this one I guess...
 
Last edited:
Back
Top Bottom