moebiusband/MD-Bench
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix Simd2xnn Kernel with half neighbor lists and add AVX512 intrinsics with double

Browse Source 
Signed-off-by: Rafael Ravedutti <rafaelravedutti@gmail.com>
This commit is contained in:
Rafael Ravedutti 2022-03-23 15:21:07 +01:00
parent 94521f03b3
commit e72323ab6a
2 changed files with 17 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
gromacs/includes/simd/avx512_double.h
View File

@ -40,6 +40,7 @@ static inline MD_SIMD_MASK simd_mask_cond_lt(MD_SIMD_FLOAT a, MD_SIMD_FLOAT b) {
static inline MD_SIMD_MASK simd_mask_from_u32(unsigned int a) { return _cvtu32_mask8(a); }
static inline unsigned int simd_mask_to_u32(MD_SIMD_MASK a) { return _cvtmask8_u32(a); }
static inline MD_SIMD_FLOAT simd_load(MD_FLOAT *p) { return _mm512_load_pd(p); }
static inline MD_SIMD_FLOAT select_by_mask(MD_SIMD_FLOAT a, MD_SIMD_MASK m) { return _mm512_mask_mov_pd(_mm512_setzero_pd(), m, a); }
static inline MD_FLOAT simd_h_reduce_sum(MD_SIMD_FLOAT a) {
MD_SIMD_FLOAT x = _mm512_add_pd(a, _mm512_shuffle_f64x2(a, a, 0xee));
x = _mm512_add_pd(x, _mm512_shuffle_f64x2(x, x, 0x11));
@ -94,3 +95,17 @@ static inline MD_FLOAT simd_h_dual_incr_reduced_sum(MD_FLOAT *m, MD_SIMD_FLOAT v
t0 = _mm512_add_pd(t0, _mm512_permutex_pd(t0, 0xb1));
return _mm_cvtsd_f64(_mm512_castpd512_pd128(t0));
}
inline void simd_h_decr(MD_FLOAT *m, MD_SIMD_FLOAT a) {
__m256d t;
a = _mm512_add_pd(a, _mm512_shuffle_f64x2(a, a, 0xee));
t = _mm256_load_pd(m);
t = _mm256_sub_pd(t, _mm512_castpd512_pd256(a));
_mm256_store_pd(m, t);
}
static inline void simd_h_decr3(MD_FLOAT *m, MD_SIMD_FLOAT a0, MD_SIMD_FLOAT a1, MD_SIMD_FLOAT a2) {
simd_h_decr(m, a0);
simd_h_decr(m + CLUSTER_N, a1);
simd_h_decr(m + CLUSTER_N * 2, a2);
}

4
gromacs/includes/simd/avx512_float.h
View File

@ -95,6 +95,6 @@ inline void simd_h_decr(MD_FLOAT *m, MD_SIMD_FLOAT a) {
static inline void simd_h_decr3(MD_FLOAT *m, MD_SIMD_FLOAT a0, MD_SIMD_FLOAT a1, MD_SIMD_FLOAT a2) {
simd_h_decr(m, a0);
simd_h_decr(m + CLUSTER_M, a1);
simd_h_decr(m + CLUSTER_M * 2, a2);
simd_h_decr(m + CLUSTER_N, a1);
simd_h_decr(m + CLUSTER_N * 2, a2);
}