tune matmul for gcn

Author: netrunnereve

ggml/src/ggml-vulkan/ggml-vulkan.cpp

@@ -1740,6 +1740,10 @@ static void ggml_vk_load_shaders(vk_device& device) {
         m_warptile_mmq_int = { 128,  64,  64, 32, subgroup_size_8,     32, 2, 2, 2, 1, subgroup_size_8 };
         s_warptile_mmq_int = { subgroup_size_32, 32, 32, 32, 32,       32, 2, 2, 1, 1, subgroup_size_8 };
 
+        // chip specific tuning
+        if (device->architecture == AMD_GCN)
+            m_warptile_mmq = { 256, 64, 64, 32, subgroup_size_8, 16, 2, 2, tn_m, tk_m, subgroup_size_8 };
+
         l_mmq_wg_denoms = l_wg_denoms = {128, 128, 1 };
         m_mmq_wg_denoms = m_wg_denoms = { 64,  64, 1 };
         s_mmq_wg_denoms = s_wg_denoms = { 32,  32, 1 };