vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

q4_0c: prefetch on AVX-512 and ARM

Browse source 
Seems significant especially for evaluation time
This commit is contained in:
Håkon H. Hitland 2023-04-24 18:17:31 +02:00
parent 1b49d26f8a
commit 2949725fea
1 changed files with 13 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

13
ggml.c
View file

@ -3126,10 +3126,17 @@ static void ggml_vec_dot_q4_0c_q8_0c(const int n, float * restrict s, const void
float sumf = 0.0; float sumf = 0.0;
#if defined(__ARM_NEON) #if defined(__ARM_NEON)
const int ahead=80;
float32x4_t sumv0 = vdupq_n_f32(0.0f); float32x4_t sumv0 = vdupq_n_f32(0.0f);
float32x4_t sumv1 = vdupq_n_f32(0.0f); float32x4_t sumv1 = vdupq_n_f32(0.0f);
for (int i = 0; i < nb/2; i++) { for (int i = 0; i < nb/2; i++) {
__builtin_prefetch(&xqs[i*QK4_0 + 64*ahead]);
__builtin_prefetch(&yqs[2*i*QK8_0C + 64*ahead]);
__builtin_prefetch(&yqs[2*i*QK8_0C + 64*ahead + 64]);
__builtin_prefetch(&xds[2*i + 64/4*ahead]);
__builtin_prefetch(&yds[2*i + 64/4*ahead]);
const int dst0 = i + i/2*2; // 0, 1, 4, 5, 8, 9, ... const int dst0 = i + i/2*2; // 0, 1, 4, 5, 8, 9, ...
const int dst1 = i + i/2*2 + 2; // 2, 3, 6, 7, 10, 11 ... const int dst1 = i + i/2*2 + 2; // 2, 3, 6, 7, 10, 11 ...
@ -3188,9 +3195,15 @@ static void ggml_vec_dot_q4_0c_q8_0c(const int n, float * restrict s, const void
sumf = vaddvq_f32(sumv0) + vaddvq_f32(sumv1); sumf = vaddvq_f32(sumv0) + vaddvq_f32(sumv1);
#elif defined(__AVX512F__) #elif defined(__AVX512F__)
const int ahead = 64;
// Initialize accumulator with zeros // Initialize accumulator with zeros
__m512 acc = _mm512_setzero_ps(); __m512 acc = _mm512_setzero_ps();
for (int i = 0; i < nb; i += 4) { for (int i = 0; i < nb; i += 4) {
_mm_prefetch(xqs + i*QK4_0/2 + 64*ahead, _MM_HINT_T0);
_mm_prefetch(yqs + i*QK8_0 + 64*ahead, _MM_HINT_T0);
_mm_prefetch(yqs + i*QK8_0 + 64*ahead + 64, _MM_HINT_T0);
_mm_prefetch(xds + i + 64/4*ahead, _MM_HINT_T0);
_mm_prefetch(yds + i + 64/4*ahead, _MM_HINT_T0);
acc = dot_q4_0c_fourblocks_avx512(acc, xqs + i*QK4_0/2, xds + i, yqs + i*QK8_0, yds + i); acc = dot_q4_0c_fourblocks_avx512(acc, xqs + i*QK4_0/2, xds + i, yqs + i*QK8_0, yds + i);
} }
// Horizontal sum of all lanes of the accumulator // Horizontal sum of all lanes of the accumulator