Another shot at AVX-512 support

Makefile

@@ -43,7 +43,7 @@ endif
 #       feel free to update the Makefile for your architecture and send a pull request or issue
 ifeq ($(UNAME_M),x86_64)
 	# AVX 512
-	CFLAGS += -mavx512f -mavx512dq -mfma -mf16c
+	CFLAGS += -mavx512f -mfma -mf16c
 
 	# AVX 256
 	#CFLAGS += -mavx -mavx2 -mfma -mf16c

ggml.c

@@ -568,37 +568,16 @@ inline static void ggml_vec_dot_f16(const int n, float * restrict s, ggml_fp16_t
     __m512 x0, x1, x2, x3;
     __m512 y0, y1, y2, y3;
 
-    __m256 t0lo, t0hi, t1lo, t1hi, t2lo, t2hi, t3lo, t3hi;
-
     for (int i = 0; i < n64; i += 64) {
-        // TODO: is this the best way to do this?
+        x0 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(x + i + 0 )));
-        t0lo = _mm256_loadu_ps((const float*)(x + i + 0 ));
+        x1 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(x + i + 16)));
-        t0hi = _mm256_loadu_ps((const float*)(x + i + 8 ));
+        x2 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(x + i + 32)));
-        t1lo = _mm256_loadu_ps((const float*)(x + i + 16));
+        x3 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(x + i + 48)));
-        t1hi = _mm256_loadu_ps((const float*)(x + i + 24));
+
-        t2lo = _mm256_loadu_ps((const float*)(x + i + 32));
+        y0 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(y + i + 0 )));
-        t2hi = _mm256_loadu_ps((const float*)(x + i + 40));
+        y1 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(y + i + 16)));
-        t3lo = _mm256_loadu_ps((const float*)(x + i + 48));
+        y2 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(y + i + 32)));
-        t3hi = _mm256_loadu_ps((const float*)(x + i + 56));
+        y3 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(y + i + 48)));
-
-        x0   = _mm512_insertf32x8(_mm512_castps256_ps512(t0lo), t0hi, 1);
-        x1   = _mm512_insertf32x8(_mm512_castps256_ps512(t1lo), t1hi, 1);
-        x2   = _mm512_insertf32x8(_mm512_castps256_ps512(t2lo), t2hi, 1);
-        x3   = _mm512_insertf32x8(_mm512_castps256_ps512(t3lo), t3hi, 1);
-
-        t0lo = _mm256_loadu_ps((const float*)(y + i + 0 ));
-        t0hi = _mm256_loadu_ps((const float*)(y + i + 8 ));
-        t1lo = _mm256_loadu_ps((const float*)(y + i + 16));
-        t1hi = _mm256_loadu_ps((const float*)(y + i + 24));
-        t2lo = _mm256_loadu_ps((const float*)(y + i + 32));
-        t2hi = _mm256_loadu_ps((const float*)(y + i + 40));
-        t3lo = _mm256_loadu_ps((const float*)(y + i + 48));
-        t3hi = _mm256_loadu_ps((const float*)(y + i + 56));
-
-        y0   = _mm512_insertf32x8(_mm512_castps256_ps512(t0lo), t0hi, 1);
-        y1   = _mm512_insertf32x8(_mm512_castps256_ps512(t1lo), t1hi, 1);
-        y2   = _mm512_insertf32x8(_mm512_castps256_ps512(t2lo), t2hi, 1);
-        y3   = _mm512_insertf32x8(_mm512_castps256_ps512(t3lo), t3hi, 1);
 
         sum0 = _mm512_fmadd_ps(x0, y0, sum0);
         sum1 = _mm512_fmadd_ps(x1, y1, sum1);
@@ -953,36 +932,26 @@ inline static void ggml_vec_mad_f16(const int n, ggml_fp16_t * restrict y, ggml_
     __m512 x0, x1, x2, x3;
     __m512 y0, y1, y2, y3;
 
-    __m256 t0lo, t0hi, t1lo, t1hi, t2lo, t2hi, t3lo, t3hi;
-
     for (int i = 0; i < n64; i += 64) {
-        t0lo = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(y + i + 0 )));
+        x0 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(x + i + 0 )));
-        t0hi = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(y + i + 8 )));
+        x1 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(x + i + 16)));
-        t1lo = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(y + i + 16)));
+        x2 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(x + i + 32)));
-        t1hi = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(y + i + 24)));
+        x3 = _mm512_cvtph_ps(_mm256_loadu_si256((__m256i*)(x + i + 48)));
-        t2lo = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(y + i + 32)));
+
-        t2hi = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(y + i + 40)));
+        y0 = _mm512_loadu_ps(y + i + 0 );
-        t3lo = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(y + i + 48)));
+        y1 = _mm512_loadu_ps(y + i + 16);
-        t3hi = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(y + i + 56)));
+        y2 = _mm512_loadu_ps(y + i + 32);
-
+        y3 = _mm512_loadu_ps(y + i + 48);
-        y0 = _mm512_insertf32x8(_mm512_castps256_ps512(t0lo), t0hi, 1);
-        y1 = _mm512_insertf32x8(_mm512_castps256_ps512(t1lo), t1hi, 1);
-        y2 = _mm512_insertf32x8(_mm512_castps256_ps512(t2lo), t2hi, 1);
-        y3 = _mm512_insertf32x8(_mm512_castps256_ps512(t3lo), t3hi, 1);
-
-        t0lo = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(x + i + 0 )));
-        t0hi = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(x + i + 8 )));
-        t1lo = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(x + i + 16)));
-        t1hi = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(x + i + 24)));
-        t2lo = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(x + i + 32)));
-        t2hi = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(x + i + 40)));
-        t3lo = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(x + i + 48)));
-        t3hi = _mm256_cvtph_ps(_mm_loadu_si128((__m128i*)(x + i + 56)));
 
         y0 = _mm512_fmadd_ps(x0, v16, y0);
         y1 = _mm512_fmadd_ps(x1, v16, y1);
         y2 = _mm512_fmadd_ps(x2, v16, y2);
         y3 = _mm512_fmadd_ps(x3, v16, y3);
+
+        _mm256_storeu_si256((__m256i*)(y + i + 0 ), _mm512_cvtps_ph(y0, 0));
+        _mm256_storeu_si256((__m256i*)(y + i + 16), _mm512_cvtps_ph(y1, 0));
+        _mm256_storeu_si256((__m256i*)(y + i + 32), _mm512_cvtps_ph(y2, 0));
+        _mm256_storeu_si256((__m256i*)(y + i + 48), _mm512_cvtps_ph(y3, 0));
     }
 
     // leftovers