Recently, I've been monitoring my boot drive's write amplification factor and noticed some unusual behavior. After using CrystalDisk, I found that my boot drive's write amounts seem disproportionate, especially when it's idling or I'm just browsing the web. For instance, the NAND write total increased by 120GB while the host write only increased by 2GB in a short timeframe, leading to concerns. I also let the computer run in BIOS for 40 minutes, which surprisingly resulted in a 600GB use of NAND capacity. Given that my drive is a Kingston A400 and observed to be low quality with a maximum write limit of 300TB, I'm worried it might die soon with this kind of usage pattern. Is this kind of write amplification normal, or could there be an underlying issue? I've cleaned up space on the drive a bit, but I'm still concerned.
2 Answers
It sounds like you're experiencing some typical behavior for QLC drives like the Kingston A400. CrystalDisk's reports can sometimes be misleading since they rely on firmware and can be reverse-engineered. These drives tend to have higher write amplification because of their architecture. When you're not actively writing data, the drive is probably busy managing and moving data internally, which can cause those high NAND writes you’re seeing. Just keep an eye on it, and consider replacing it when it reaches that 300TB limit. It is low quality and can fail faster than expected.
The drive being over 80% full often leads to increased write amplification, especially during garbage collection processes. It sounds like the accumulation of data could have caused the weird jumps in reports. You might want to use a higher-quality NVMe drive as your boot drive in the future. They can manage those write operations much better. Now that you’ve cleaned some space, it should perform more optimally. Just keep monitoring it regularly for any unexpected changes!
Thanks for the suggestion! I didn’t consider that about the NVMe. I’ll definitely look into upgrading once my current one gives up the ghost.
Gotcha! That makes sense, especially considering it’s a QLC drive. I had a hunch that it was managing data on its own. Thanks for clarifying!