![]() So if you're writing at 500 MB/s it will take over 2.5 minutes to exhaust the cache, but since the SLC isn't being saturated and the P1 is aggressive with writes you're actually freeing up 80 MB/s in the background, so the cache is effectively 12GB larger. By default it will have 76GB of SLC, write at 2000 MB/s (SLC) and fold at 80 MB/s (QLC). The dynamic cache can also increase wear because you'll be writing/erasing TLC blocks effectively twice. So you have less OP which can increase write amplification (NAND wear). Additionally when the drive is 75%+ filled it has only static SLC left, but as mentioned this is reserved space - 48GiB of QLC is required for 12GiB of SLC. Why would you do this? To improve endurance - folding writes out sequentially. So folding here means all incoming data must go through the SLC cache first. This is the compression of SLC blocks into TLC blocks - it has a performance penalty and is slow in general. The third tier is folding which is up to 1/2 the speed of the native flash. ![]() There are reasons for this I'll get into briefly below. The second tier on most drives is direct-to-NAND, e.g.While dynamic shifts between QLC and SLC, the static is in reserved space for the lifetime of the device and thus has far higher endurance. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |