micro-optimization-CodePudding

Tags > micro-optimization

08-28front endLEA vs MOV INC for incrementing a variable and storing it in another one?
08-25Software designlea or mov inc for incrementing a variable and storing it in another one?
08-24Software designWhat's the most efficient way to swap 4 16-bit integers on a 64-bit processor?
07-26MobileFastest way to take the average of two signed integers in x86 assembly?
07-25Software designEfficiently loading both RAX and R8 with the same small positive number
07-20Enterprise`test` vs `cmp` for one-bit registers comparison
06-27OSWhy does gcc implement fmin and fmax in three different ways?
06-08EnterpriseIs there a better way to any detect bits that are set in a 16-byte array of flags?
05-28Software engineeringImplement a function that rotates 90 degrees in a two-dimensional array that represents a color imag
05-20MobileHow do you reason about fluctuations in benchmarking data?
05-20BlockchainOptimize lookup tables to simple ALU
04-30BlockchainFastest way to set highest order bit of rax register to lowest order bit in rdx register
03-26EnterpriseIs it still worth using the Quake fast inverse square root algorithm nowadays on x86-64?
03-02OSCopy bit of one register to another register (x86-64 asm)
12-21BlockchainGEMM kernel implemented using AVX2 is faster than AVX2/FMA on a Zen 2 CPU
12-21OSGEMM kernel implemented using AVX2 is faster than AVX2/FMA on a Zen 2 CPU
12-14BlockchainIntel JCC Erratum - what is the effect of prefixes used for mitigation?
12-01Software engineeringDo 32-bit and 64-bit registers cause differences in CPU micro architecture?
11-30Mobilestring_view Vs const char* performance
11-16OSWhen joining four 1-byte vars into one 4-byte word, which is a faster way to shift and OR ? (compari
11-14front endHow to write a custom exception class derived from std::invalid_argument?
11-06front endC different using declarations for different concepts
11-06front endC 20 Likely and UnLikely?
10-08EnterpriseWhen source registers in avx instruction can be reused
09-30otherWhy does clang's epilogue use `add $N, %rsp` instead of `mov %rbp, %rsp` to restore `%rsp`?
09-22OSx86_64 best way to reduce 64 bit register to 32 bit retaining zero or non-zero status