3 ms·
And if you're designing hardware (or on an FPGA) ChaCha (and BLAKE2/3) are actually really freakin' fast for the die area they take. IIRC ChaCha20 beats AES in
by SAI_Peregrinus 5y ago
And if you're designing hardware (or on an FPGA) ChaCha (and BLAKE2/3) are actually really freakin' fast for the die area they take. IIRC ChaCha20 beats AES in hardware. It's just extremely rare (FPGA only) to find it in hardware, since there's not enough demand. I expect that to eventually change.
- api 5y agoGiven that all these are ARX cores, I wonder if a fused ARX instruction could cover a wide range of them?
- SAI_Peregrinus 5y agoProbably. With a barrel shifter easily. Barrel shifters are a bit slower than wired shifts though, so for ultimate speed you'd end up with hardware shift amounts.
- api 5y agoThe advantage would be a single instruction that implemented the core of a wide range of things including BLAKE2, BLAKE3, Salsa, ChaCha, Speck, and more.
- brandmeyer 5y agoARMv8.2 has rotate-and-xor and xor-and-rotate so that the (extremely cheap) xor can be saved.