Skip to content

Instantly share code, notes, and snippets.

View magurosan's full-sized avatar

Masaki Ota magurosan

  • Nagoya, Aichi, Japan
View GitHub Profile
class inst l/t cpi ipc
reg64 add latency 3.633499e-01 2.752168e+00
reg64 add throughput 7.145596e-02 1.399463e+01
reg64 lea latency 1.119077e-02 8.935934e+01
reg64 lea throughput 1.119077e-02 8.935934e+01
reg64 xor dst,dst latency 4.510581e-02 2.217009e+01
reg64 xor dst,dst throughput 4.510581e-02 2.217009e+01
reg64 xor latency 4.510093e-02 2.217249e+01
reg64 xor throughput 4.511082e-02 2.216763e+01
reg64 load latency 9.582788e-01 1.043538e+00
@magurosan
magurosan / xoshiro256_x8_avx512.c
Created May 26, 2022 14:40
x8 xoshiro256 (AVX-512)
#include <immintrin.h>
#include <stdint.h>
union PARALLEL_XOSHIRO_AVX512_STATE {
__m512i state512[4];
uint64_t state64[32];
uint32_t state32[64];
};
typedef union PARALLEL_XOSHIRO_AVX512_STATE xoshiro256_x8_avx512_state_t;