4 ms·

What do you mean? Digital storage is kind of pointless if we cannot know exactly which bit came from what logical location. ie: it's not like analog audio tap

by LASR 2y ago

What do you mean? Digital storage is kind of pointless if we cannot know exactly which bit came from what logical location. ie: it's not like analog audio tape, where we get some signal, but don't really know its exact position.

yjftsjthsd-h 2y ago
I believe it's that you can't read that small an amount at once; AIUI most memory will only let you read to the byte at most and then you'd have to do some bit-fiddling to pull out individual bits, and I vaguely recall that flash is often only read/write-able in like 4k blocks at a time, so if you want to know the first bit on the device what you actually do is copy the first 4kB into RAM then grab the first byte of the result and AND it with 1<<64 to isolate just the first bit
dessimus 2y ago
Because we don't request specific bits from Storage, we address by block, which returns 4k of bits (or whatever the formatted block size is), and then process the data within that block.
cptskippy 2y ago
Well I was a little wrong, modern NAND supports page, word, and bit level addressing. Older NOR flash only support page and word level addressing. But to access that 1 bit, the flash controller is still reading multiple bits. With all flash, a cell is the smallest addressable unit. Bits are stored in cells and depending on the flash type it can store anywhere from 1 to 5 bits per cell. So with SLC flash a cell is analogous to a bit, but for the multi-cell flash types a cell might mean 2, 3, 4 or 5 bits. Flash cells are organized into strings, pages, blocks, planes, and then the flash die itself. Reading and writing of NAND flash occurs at the page level (4-16 KiB per page) but erasing occurs at the block level. So in the case of NAND, the controller cannot read/write a single bit of data. So while you can specify that only 1 bit be changed in flash storage, there's an entire abstraction layer between you and the flash chip that hides the fact that a lot of data gets moved around to support a single bit being changed.

by LASR 2y ago

yjftsjthsd-h 2y ago

I believe it's that you can't read that small an amount at once; AIUI most memory will only let you read to the byte at most and then you'd have to do some bit-fiddling to pull out individual bits, and I vaguely recall that flash is often only read/write-able in like 4k blocks at a time, so if you want to know the first bit on the device what you actually do is copy the first 4kB into RAM then grab the first byte of the result and AND it with 1<<64 to isolate just the first bit

dessimus 2y ago

Because we don't request specific bits from Storage, we address by block, which returns 4k of bits (or whatever the formatted block size is), and then process the data within that block.

cptskippy 2y ago

Well I was a little wrong, modern NAND supports page, word, and bit level addressing. Older NOR flash only support page and word level addressing. But to access that 1 bit, the flash controller is still reading multiple bits. With all flash, a cell is the smallest addressable unit. Bits are stored in cells and depending on the flash type it can store anywhere from 1 to 5 bits per cell. So with SLC flash a cell is analogous to a bit, but for the multi-cell flash types a cell might mean 2, 3, 4 or 5 bits. Flash cells are organized into strings, pages, blocks, planes, and then the flash die itself. Reading and writing of NAND flash occurs at the page level (4-16 KiB per page) but erasing occurs at the block level. So in the case of NAND, the controller cannot read/write a single bit of data. So while you can specify that only 1 bit be changed in flash storage, there's an entire abstraction layer between you and the flash chip that hides the fact that a lot of data gets moved around to support a single bit being changed.