5 ms·
Is there hope the technique can be accelerated to practical speeds at all?
by sophistication 8y ago
Is there hope the technique can be accelerated to practical speeds at all?
- chii 8y agoOr that a stray cosmic ray won't disrupt the molecules and cause read corruptions?
- fooker 8y agoThat's not a big deal. With this density, you can just duplicate (or more) the data and still have an useful capacity. Parity based methods would do better, I guess.
- rbanffy 8y agoDuplication halves the write speed (and only tell you the data is bad, at half the read speed). ECC would reduce RW speed somewhat still, but, at least, would allow one to correct the data and rewrite it.
- fooker 8y agoDuplication does not have to be online, especially if speed is a concern.
- eximius 8y agoDuplication is a form of ECC (admittedly a bad one). If I write everything 3 times, I can correct from any single error, but I will erroneously correct on two errors. If you're limited to duplication, you'd want four copies for 1 error correction and 2 error detection.
- rbanffy 8y ago> 3 times That would be triplication. ;-) Duplication is only single error detection.
- waynecochran 8y agoMore critical than cosmic rays, the article mentions there are errors in reading now: “Sometimes hydrogen atoms can be removed from incorrect locations.”
- namibj 8y agoSounds like DVD-RW...
- nine_k 8y agoMaybe reading can be made fast(er); a very compact ROM can be pretty useful. Much more compact than a bluray disc, for instance.