iq1bn: faster scalar dot product

At the end of the day, lookup is still better when not using simd.
This scalar dot product version gets us 14.7 t/s on a Ryzen-7950X
with 16 threads (up from 10.5 t/s).

1 files changed, 47 insertions, 3 deletions

diff --git a/iqk-quantize.cpp b/iqk-quantize.cpp
index b2100ac4..c7b868a5 100644
--- a/iqk-quantize.cpp
+++ b/iqk-quantize.cpp
@@ -191,6 +191,49 @@ inline int8_t iq1bn_dequant(uint8_t q, int i) {
 }
 }
 
+static const int8_t iq1bn_values[1280] = {
+    -1, -1, -1, -1, -1,  0,  0,  0,  0,  0,  0, -1, -1, -1, -1,  1, -1, -1, -1, -1, -1,  0, -1, -1, -1,  0,  0, -1, -1, -1,  1,  0,
+    -1, -1, -1, -1,  1, -1, -1, -1,  0,  1, -1, -1, -1,  1,  1, -1, -1, -1, -1, -1,  0, -1, -1,  0, -1,  0, -1, -1,  1, -1,  0, -1,
+    -1, -1,  0,  0, -1, -1,  0,  0,  0, -1, -1,  1,  0,  0, -1, -1, -1,  1,  0, -1, -1,  0,  1,  0, -1, -1,  1,  1,  0, -1, -1, -1,
+    -1,  1, -1, -1,  0,  0,  0,  0,  0,  0, -1,  1, -1, -1,  1, -1,  1, -1, -1, -1,  0,  1, -1, -1,  0,  0,  1, -1, -1,  1,  0,  1,
+    -1, -1, -1,  1,  1, -1, -1,  0,  1,  1, -1, -1,  1,  1,  1, -1, -1, -1, -1, -1,  0, -1,  0, -1, -1,  0, -1,  1, -1, -1,  0, -1,
+    -1,  0, -1,  0, -1,  0,  0, -1,  0, -1,  1,  0, -1,  0, -1, -1,  1, -1,  0, -1,  0,  1, -1,  0, -1,  1,  1, -1,  0, -1, -1, -1,
+     0,  0, -1,  0, -1,  0,  0, -1,  0,  0,  0,  0,  0,  1, -1,  0,  0, -1, -1,  0,  0,  0, -1,  0,  0,  0,  0, -1,  1,  0,  0,  0,
+    -1, -1,  1,  0,  0, -1,  0,  1,  0,  0, -1,  1,  1,  0,  0, -1, -1, -1,  1,  0, -1,  0, -1,  1,  0, -1,  1, -1,  1,  0, -1, -1,
+     0,  1,  0, -1,  0,  0,  1,  0, -1,  1,  0,  1,  0, -1, -1,  1,  1,  0, -1,  0,  1,  1,  0, -1,  1,  1,  1,  0, -1, -1, -1, -1,
+     1, -1,  0, -1, -1,  1, -1,  1, -1, -1,  1, -1,  0,  0,  0,  0,  0, -1,  0, -1,  1, -1,  0,  0, -1,  1, -1,  1,  0, -1,  1, -1,
+    -1,  1, -1,  1, -1,  0,  1, -1,  1, -1,  1,  1, -1,  1, -1, -1, -1,  0,  1, -1,  0, -1,  0,  1, -1,  1, -1,  0,  1, -1, -1,  0,
+     0,  1, -1,  0,  0,  0,  1, -1,  1,  0,  0,  1, -1, -1,  1,  0,  1, -1,  0,  1,  0,  1, -1,  1,  1,  0,  1, -1, -1, -1,  1,  1,
+    -1,  0, -1,  1,  1, -1,  1, -1,  1,  1, -1,  0,  0,  0,  0,  0, -1,  0,  1,  1, -1,  0,  0,  1,  1, -1,  1,  0,  1,  1, -1, -1,
+     1,  1,  1, -1,  0,  1,  1,  1, -1,  1,  1,  1,  1, -1, -1, -1, -1, -1,  0,  0, -1, -1, -1,  0,  1, -1, -1, -1,  0, -1,  0, -1,
+    -1,  0,  0,  0, -1, -1,  0,  1,  0, -1, -1,  0, -1,  1, -1, -1,  0,  0,  1, -1, -1,  0,  1,  1, -1, -1,  0, -1, -1,  0, -1,  0,
+     0, -1,  0, -1,  0,  1, -1,  0, -1,  0, -1,  0,  0, -1,  0,  0,  0,  0,  0,  0,  0,  0,  0, -1,  0,  1,  0,  0, -1,  0, -1,  1,
+     0, -1,  0,  0,  1,  0, -1,  0,  1,  1,  0, -1,  0, -1, -1,  1, -1,  0,  0, -1,  1, -1,  0,  1, -1,  1, -1,  0, -1,  0,  1, -1,
+     0,  0,  0,  1, -1,  0,  1,  0,  1, -1,  0, -1,  1,  1, -1,  0,  0,  1,  1, -1,  0,  1,  1,  1, -1,  0, -1, -1, -1,  0,  0,  0,
+    -1, -1,  0,  0,  1, -1, -1,  0,  0, -1,  0, -1,  0,  0,  0,  0, -1,  0,  0,  0,  0,  0,  0,  0,  1,  0, -1,  0,  0, -1,  1, -1,
+     0,  0,  0,  1, -1,  0,  0,  1,  1, -1,  0,  0, -1, -1,  0,  0,  0,  0, -1,  0,  0,  0,  1, -1,  0,  0,  0, -1,  0,  0,  0,  0,
+     0,  0,  0,  0,  0,  1,  0,  0,  0,  0, -1,  1,  0,  0,  0,  0,  1,  0,  0,  0,  1,  1,  0,  0,  0, -1, -1,  1,  0,  0,  0, -1,
+     1,  0,  0,  1, -1,  1,  0,  0, -1,  0,  1,  0,  0,  0,  0,  1,  0,  0,  0,  0,  0,  0,  0,  1,  0,  1,  0,  0, -1,  1,  1,  0,
+     0,  0,  1,  1,  0,  0,  1,  1,  1,  0,  0, -1, -1, -1,  1,  0,  0, -1, -1,  1,  0,  1, -1, -1,  1,  0, -1,  0, -1,  1,  0,  0,
+     0, -1,  1,  0,  1,  0, -1,  1,  0, -1,  1, -1,  1,  0,  0,  1, -1,  1,  0,  1,  1, -1,  1,  0, -1, -1,  0,  1,  0,  0, -1,  0,
+     1,  0,  1, -1,  0,  1,  0, -1,  0,  0,  1,  0,  0,  0,  0,  1,  0,  1,  0,  0,  1,  0,  0,  0,  0,  0,  0, -1,  1,  0,  1,  0,
+     0,  1,  0,  1,  0,  1,  1,  0,  1,  0, -1, -1,  1,  1,  0,  0, -1,  1,  1,  0,  1, -1,  1,  1,  0, -1,  0,  1,  1,  0,  0,  0,
+     1,  1,  0,  1,  0,  1,  1,  0, -1,  1,  1,  1,  0,  0,  1,  1,  1,  0,  1,  1,  1,  1,  0, -1, -1, -1, -1,  1,  0, -1, -1, -1,
+     1,  1, -1, -1, -1,  1, -1,  0, -1, -1,  1,  0,  0, -1, -1,  1,  1,  0, -1, -1,  1, -1,  1, -1, -1,  1,  0,  0,  0,  0,  0,  0,
+     1, -1, -1,  1,  1,  1, -1, -1,  1, -1, -1,  0, -1,  1,  0, -1,  0, -1,  1,  1, -1,  0, -1,  1, -1,  0,  0, -1,  1,  0,  0,  0,
+    -1,  1,  1,  0,  0, -1,  1, -1,  1,  0, -1,  1,  0,  1,  0, -1,  1,  1,  1,  0, -1,  1, -1, -1,  1, -1,  1,  0, -1,  1, -1,  1,
+     1, -1,  1, -1,  1, -1,  0,  1, -1,  1,  0,  0,  1, -1,  1,  1,  0,  1, -1,  1, -1,  1,  1, -1,  1,  0,  0,  0,  0,  0,  0,  1,
+     1, -1,  1,  1,  1,  1, -1,  1, -1, -1, -1,  0,  1,  0, -1, -1,  0,  1,  1, -1, -1,  0,  1, -1,  0, -1,  0,  1,  0,  0, -1,  0,
+     1,  1,  0, -1,  0,  1, -1,  1, -1,  0,  1,  0,  1, -1,  0,  1,  1,  1, -1,  0,  1, -1, -1,  0,  0,  1,  0, -1,  0,  0,  1,  1,
+    -1,  0,  0,  1, -1,  0,  0,  0,  1,  0,  0,  0,  0,  1,  1,  0,  0,  0,  1, -1,  1,  0,  0,  1,  0,  1,  0,  0,  1,  0,  0,  0,
+     0,  0,  1,  1,  0,  0,  1, -1, -1,  1,  0,  1,  0, -1,  1,  0,  1,  1, -1,  1,  0,  1, -1,  0,  1,  0,  1,  0,  0,  1,  0,  1,
+     1,  0,  1,  0,  1, -1,  1,  1,  0,  1,  0,  1,  1,  0,  1,  1,  1,  1,  0,  1, -1, -1, -1,  1,  1,  0, -1, -1,  1,  1,  1, -1,
+    -1,  1,  1, -1,  0, -1,  1,  1,  0,  0, -1,  1,  1,  1,  0, -1,  1,  1, -1,  1, -1,  1,  1,  0,  1, -1,  1,  1,  1,  1, -1,  1,
+     1,  0,  0,  0,  0,  0, -1, -1,  0,  1,  1,  0, -1,  0,  1,  1,  1, -1,  0,  1,  1, -1,  0,  0,  1,  1,  0,  0,  0,  1,  1,  1,
+     0,  0,  1,  1, -1,  1,  0,  1,  1,  0,  1,  0,  1,  1,  1,  1,  0,  1,  1, -1, -1,  1,  1,  1,  0, -1,  1,  1,  1,  1, -1,  1,
+     1,  1, -1,  0,  1,  1,  1,  0,  0,  1,  1,  1,  1,  0,  1,  1,  1, -1,  1,  1,  1,  1,  0,  1,  1,  1,  1,  1,  1,  1,  1,  1,
+};
+
 void ggml_vec_dot_iq1_bn_q8_K64(int n, float * s, size_t bs, const void * vx, size_t bx, const void * vy, size_t by, int nrc) {
 
     GGML_UNUSED(bs);
@@ -218,13 +261,14 @@ void ggml_vec_dot_iq1_bn_q8_K64(int n, float * s, size_t bs, const void * vx, si
         int nb = std::min(ii + 32, nblock);
         for (int i = ii; i < nb; ++i) {
             auto ql = x[i].ql;
-            auto extra = x[i].extra;
+            const int8_t * extra = iq1bn_values + 5*x[i].extra;
             for (int i16 = 0; i16 < QK_IQ1BN/16; ++i16) {
                 for (int k = 0; k < 3; ++k) {
                     uint8_t q = *ql++;
-                    for (int j = 0; j < 5; ++j) q1[5*k+j] = iq1bn_dequant(q, j);
+                    const int8_t * vs = iq1bn_values + 5*q;
+                    for (int j = 0; j < 5; ++j) q1[5*k+j] = vs[j];
                 }
-                q1[15] = iq1bn_dequant(extra, i16);
+                q1[15] = extra[i16];
                 // We collect 8 q8 values per block into each element of sum16
                 // => 32 x 8 = 256 values in each loop over i, so this cannot overflow the int16_t range
                 //    (q8 is in -127...127, and hence the sum is in -32512...32512