Memory, The book is on the shelf in the room across the hall. 元 cache, the book is on the shelf behind you. L2 cache, the book is on your desk, but you have to find the page. L1 cache, the book is on your desk, open to the right page.
I explain anymore that CPU's getting data is like reading from a book, from the CPU's reference at least. These days, the computers always have to wait for the disk, even M2s right on the motherboard. So you'd put it every other sector so the system could keep up with the disk. While this action is quite quick, it does add up.īack in my early computer days, you'd put data in every other, or every third sector (I think they called it interleaving, or something) This is because at the time, the drive was able to read from consecutive sectors faster than the data could be processed. (edit: there is a tiny difference, but not as compared to in spinning platter drives) In a traditional HD, if the sectors are in sequence, the head streams across them, if they're spread out, the disk needs to spin and the head to move to catch the correct sector. There's no point to it since there is no minimal temporal cost in grabbing from random sectors.
0 kommentar(er)
