3 ms·
The fact that we have AES-NI (despite the fact that AES constructions are being replaced with EC constructions in quite a few protocols, due to the difficulty o
by cyphar 8y ago
The fact that we have AES-NI (despite the fact that AES constructions are being replaced with EC constructions in quite a few protocols, due to the difficulty of using AES in a safe way) should be clear that CPU designers don't really care about longevity, all that matters is current performance characteristics.
Not to mention that I think it's fair to say that UTF-8 will outlast AES (it's actually older than AES). After all, AES was only standardised in 2001 -- and AES-NI was added when it was 7 years old. Unicode (and UTF-8) were standardised in 1991. So if we have AES instructions we should already have UTF-8 instructions.
- rauhl 8y ago> The fact that we have AES-NI (despite the fact that AES constructions are being replaced with EC constructions in quite a few protocols, due to the difficulty of using AES in a safe way) Say what? Unless you mean something other than ‘elliptic curve’ by ‘EC,’ that doesn’t make sense to me: AES & EC are completely different kinds of thing, the former being symmetric & the latter being asymmetric. Indeed, it’s quite common to use an EC-derived secret as the key for AES-encrypted material.