3 ms·
Huh? Double precision SIMD? SSE2 instructions aren't slower than SSE.
by jedbrown 16y ago
Huh? Double precision SIMD? SSE2 instructions aren't slower than SSE.
- astrange 16y agoNo, single precision, as in "the fastest way to multiply a double is to use floats instead". I've seen more than enough C programs unnecessarily using double that I wish floating point constants would just be float by default. (Although it seems like clang does something special here, as I don't see many double-to-float conversions in its output.)
- jedbrown 16y agoThen you aren't multiplying doubles, it would be a shame to get a completely wrong answer on an ill-conditioned problem because the implementation disregarded the spec, cf. "the fastest way to process UTF-8 is to truncate it to ASCII", "the fastest RSA encryption is ROT-13". Also note that floats are only faster with packed instructions (often not possible without changing APIs) or when memory bandwidth is an issue.