Speaking of hash tables, I'm interested in optimizing hash tables that are usually really small. [..]
Because you’re executing log(8) == 3 cmp+branch fused insns vs. 4 slower AVX instructions.
-
-
I should test.
-
Go for it, I’m curious :)
- 2 more replies
New conversation -
-
-
I'd expect the branch predictor to make that horrible but I guess cmov could be used instead
-
but even that makes the address of later loads depend on earlier, so I bet it comes out slower
End of conversation
New conversation -
-
-
More often than not just thinking about problems differently leads me to a simple scalar solution that beats my SIMD ideas :(
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.