Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
kwillets
on Sept 8, 2017
|
parent
|
context
|
favorite
| on:
Implementing strcmp, strlen, and strstr using SSE ...
It's 16/3 bytes per cycle.
The newer SIMD stuff seems to be factoring it out into bytewise compare and CLZ on a mask.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
The newer SIMD stuff seems to be factoring it out into bytewise compare and CLZ on a mask.