vulkan: Dynamic subgroup size support for Q6_K mat_vec (llama/10536)

* subgroup 64 version with subgroup add. 15% faster

scalable version

tested for subgroup sizes 16-128

* check for subgroup multiple of 16 and greater than 16

* subgroup sizes are always a power of 2 (https://github.com/KhronosGroup/GLSL/issues/45)

* force 16 sequential threads per block

* make 16 subgroup size a constant

diff --git a/ggml/src/ggml-vulkan/ggml-vulkan.cpp b/ggml/src/ggml-vulkan/ggml-vulkan.cpp
index 849c119237bba3717028db8a7cb0e3308a785f91..df6a659f44ff5b17bdf188fce1758073effbbf43 100644 (file)
--- a/ggml/src/ggml-vulkan/ggml-vulkan.cpp
+++ b/ggml/src/ggml-vulkan/ggml-vulkan.cpp
@@ -1231,6 +1231,9 @@ static void ggml_vk_load_shaders(vk_device& device) {
 
     std::cerr << "ggml_vulkan: Compiling shaders";
 
+    // some shaders require the subgroup size to be 16 or larger
+    const uint32_t subgroup_size_16 = std::max(device->subgroup_size, 16u);
+
     // mulmat
     std::vector<uint32_t> l_warptile, m_warptile, s_warptile,
                           l_warptile_mmq, m_warptile_mmq, s_warptile_mmq;
@@ -1240,11 +1243,11 @@ static void ggml_vk_load_shaders(vk_device& device) {
 
     l_warptile = { 128, 128, 128, 16, device->subgroup_size * 2, 64, 2, 4, 4, device->subgroup_size };
     m_warptile = { 128,  64,  64, 16, device->subgroup_size, 32, 2, 4, 2, device->subgroup_size };
-    s_warptile = { std::max(device->subgroup_size, 16u),  32,  32, 16, 32, 32, 2, 2, 2, device->subgroup_size };
+    s_warptile = { subgroup_size_16,  32,  32, 16, 32, 32, 2, 2, 2, device->subgroup_size };
 
     l_warptile_mmq = { 128, 128, 128, 32, device->subgroup_size * 2, 64, 2, 4, 4, device->subgroup_size };
     m_warptile_mmq = { 128,  64,  64, 32, device->subgroup_size, 32, 2, 4, 2, device->subgroup_size };
-    s_warptile_mmq = { std::max(device->subgroup_size, 16u),  32,  32, 32, 32, 32, 2, 2, 2, device->subgroup_size };
+    s_warptile_mmq = { subgroup_size_16,  32,  32, 32, 32, 32, 2, 2, 2, device->subgroup_size };
 
     l_mmq_wg_denoms = l_wg_denoms = {128, 128, 1 };
     m_mmq_wg_denoms = m_wg_denoms = { 64,  64, 1 };
@@ -1431,7 +1434,7 @@ static void ggml_vk_load_shaders(vk_device& device) {
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q3_K], "mul_mat_vec_q3_k_f32_f32", mul_mat_vec_q3_k_f32_f32_len, mul_mat_vec_q3_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q4_K], "mul_mat_vec_q4_k_f32_f32", mul_mat_vec_q4_k_f32_f32_len, mul_mat_vec_q4_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q5_K], "mul_mat_vec_q5_k_f32_f32", mul_mat_vec_q5_k_f32_f32_len, mul_mat_vec_q5_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
-    ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q6_K], "mul_mat_vec_q6_k_f32_f32", mul_mat_vec_q6_k_f32_f32_len, mul_mat_vec_q6_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
+    ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_Q6_K], "mul_mat_vec_q6_k_f32_f32", mul_mat_vec_q6_k_f32_f32_len, mul_mat_vec_q6_k_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true);
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f32_f32[GGML_TYPE_IQ4_NL], "mul_mat_vec_iq4_nl_f32_f32", mul_mat_vec_iq4_nl_f32_f32_len, mul_mat_vec_iq4_nl_f32_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1, true);
 
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_F32 ], "mul_mat_vec_f32_f16_f32",  mul_mat_vec_f32_f16_f32_len,  mul_mat_vec_f32_f16_f32_data,  "main", 3, sizeof(vk_mat_vec_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1);
@@ -1445,7 +1448,7 @@ static void ggml_vk_load_shaders(vk_device& device) {
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q3_K], "mul_mat_vec_q3_k_f16_f32", mul_mat_vec_q3_k_f16_f32_len, mul_mat_vec_q3_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q4_K], "mul_mat_vec_q4_k_f16_f32", mul_mat_vec_q4_k_f16_f32_len, mul_mat_vec_q4_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q5_K], "mul_mat_vec_q5_k_f16_f32", mul_mat_vec_q5_k_f16_f32_len, mul_mat_vec_q5_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
-    ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q6_K], "mul_mat_vec_q6_k_f16_f32", mul_mat_vec_q6_k_f16_f32_len, mul_mat_vec_q6_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
+    ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_Q6_K], "mul_mat_vec_q6_k_f16_f32", mul_mat_vec_q6_k_f16_f32_len, mul_mat_vec_q6_k_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true);
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_f16_f32[GGML_TYPE_IQ4_NL], "mul_mat_vec_iq4_nl_f16_f32", mul_mat_vec_iq4_nl_f16_f32_len, mul_mat_vec_iq4_nl_f16_f32_data, "main", 3, sizeof(vk_mat_vec_push_constants), {2, 1, 1}, {device->subgroup_size}, 1, true);
 
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_F32 ], "mul_mat_vec_id_f32_f32",  mul_mat_vec_id_f32_f32_len,  mul_mat_vec_id_f32_f32_data,  "main", 4, sizeof(vk_mat_vec_id_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1);
@@ -1459,7 +1462,7 @@ static void ggml_vk_load_shaders(vk_device& device) {
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q3_K], "mul_mat_vec_id_q3_k_f32", mul_mat_vec_id_q3_k_f32_len, mul_mat_vec_id_q3_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q4_K], "mul_mat_vec_id_q4_k_f32", mul_mat_vec_id_q4_k_f32_len, mul_mat_vec_id_q4_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q5_K], "mul_mat_vec_id_q5_k_f32", mul_mat_vec_id_q5_k_f32_len, mul_mat_vec_id_q5_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
-    ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q6_K], "mul_mat_vec_id_q6_k_f32", mul_mat_vec_id_q6_k_f32_len, mul_mat_vec_id_q6_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1, 1, 1}, {device->subgroup_size}, 1, true);
+    ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_Q6_K], "mul_mat_vec_id_q6_k_f32", mul_mat_vec_id_q6_k_f32_len, mul_mat_vec_id_q6_k_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {1, 1, 1}, {subgroup_size_16}, 1, true);
     ggml_vk_create_pipeline(device, device->pipeline_dequant_mul_mat_vec_id_f32[GGML_TYPE_IQ4_NL], "mul_mat_vec_id_iq4_nl_f32", mul_mat_vec_id_iq4_nl_f32_len, mul_mat_vec_id_iq4_nl_f32_data, "main", 4, sizeof(vk_mat_vec_id_push_constants), {2, 1, 1}, {device->subgroup_size, 2}, 1, true);
 
     // dequant shaders

diff --git a/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q6_k.comp b/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q6_k.comp
index 0b392d68d04d0a8b4e5a323790304499caa39a21..760aff85499f40e720f470b8f73d527bd5dec918 100644 (file)
--- a/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q6_k.comp
+++ b/ggml/src/ggml-vulkan/vulkan-shaders/mul_mat_vec_q6_k.comp
@@ -4,9 +4,11 @@
 
 #include "mul_mat_vec_base.comp"
 
-layout(local_size_x = 32, local_size_y = 1, local_size_z = 1) in;
+layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
 
-shared FLOAT_TYPE tmp[32];
+layout (constant_id = 0) const uint BLOCK_SIZE = 32;
+
+shared FLOAT_TYPE tmp[BLOCK_SIZE];
 
 void main() {
     const uint row = gl_WorkGroupID.x + gl_NumWorkGroups.x * gl_WorkGroupID.z;
@@ -21,21 +23,19 @@ void main() {
     const uint num_blocks_per_row = p.ncols / QUANT_K;
     const uint ib0 = a_offset / QUANT_K + row*num_blocks_per_row;
 
-    const uint tid = gl_LocalInvocationID.x/K_QUANTS_PER_ITERATION;  // 0...31 or 0...16
-    const uint ix  = gl_LocalInvocationID.x%K_QUANTS_PER_ITERATION;  // 0 or 0, 1
+    // 16 threads are used to process each block
+    const uint it_size = gl_WorkGroupSize.x/16;
+    const uint tid = gl_LocalInvocationID.x;
+    const uint itid = tid%16;  // 0...16
+    const uint ix  = tid/16;
 
-    const uint step = 16/K_QUANTS_PER_ITERATION;            // 16 or 8
+    const uint step = 8;
 
-    const uint v_im = tid/step;                             // 0 or 1. 0 computes 0..., 1 computes 128...
-    const uint v_in = tid - step*v_im;                      // 0...15 or 0...7
+    const uint v_im = itid/step;                            // 0 or 1. 0 computes 0..., 1 computes 128...
+    const uint v_in = itid - step*v_im;                     // 0...15 or 0...7
 
-#if K_QUANTS_PER_ITERATION == 1
-    const uint l0 = v_in;                                   // 0...15
-    const uint is = 0;
-#else
     const uint l0 = 4 * v_in;                               // 0, 4, 8, ..., 28
     const uint is = v_in / 4;
-#endif
 
     const uint ql_offset = 64*v_im + l0;
     const uint qh_offset = 32*v_im + l0;
@@ -44,7 +44,7 @@ void main() {
 
     FLOAT_TYPE temp = FLOAT_TYPE(0.0); // partial sum for thread in warp
 
-    [[unroll]] for (uint i = ix; i < num_blocks_per_row; i += K_QUANTS_PER_ITERATION) {
+    [[unroll]] for (uint i = ix; i < num_blocks_per_row; i += it_size) {
         const uint y_idx   = i * QUANT_K + y_offset;
 
         const FLOAT_TYPE d = FLOAT_TYPE(data_a[ib0 + i].d);
@@ -95,10 +95,10 @@ void main() {
     }
 
     tmp[gl_LocalInvocationID.x] = temp;
-
     // sum up partial sums and write back result
+
     barrier();
-    [[unroll]] for (uint s = 16; s > 0; s >>= 1) {
+    [[unroll]] for (uint s = gl_WorkGroupSize.x/2; s > 0; s >>= 1) {
         if (tid < s) {
             tmp[tid] += tmp[tid + s];
         }