static mode

host/driver_offline/include/device_convolution_forward_implicit_gemm_v5r1_dlops_nc0hwc1_kc0yxc1_nk0hwk1.hpp

@@ -61,6 +61,13 @@ void device_convolution_forward_implicit_gemm_v5r1_dlops_nc0hwc1_kc0yxc1_nk0hwk1
     in_n_c0_hi_wi_c1_device_buf.ToDevice(in_n_c0_hi_wi_c1.mData.data());
     wei_k_c0_y_x_c1_device_buf.ToDevice(wei_k_c0_y_x_c1.mData.data());
 
+    constexpr index_t InWeiVectorSize = 8;
+
+    if(C1 % InWeiVectorSize != 0)
+    {
+        throw std::runtime_error("wrong! C1 cannot be divided by InWeiVectorSize");
+    }
+
 #if 0
     constexpr index_t BlockSize = 256;
 
@@ -115,15 +122,17 @@ void device_convolution_forward_implicit_gemm_v5r1_dlops_nc0hwc1_kc0yxc1_nk0hwk1
     constexpr index_t CThreadTransferDstScalarPerVector_K = 8;
 #endif
 
-    constexpr index_t InWeiVectorSize = 8;
-
     const auto in_n_c0_hi_wi_c1_desc =
-        make_naive_tensor_descriptor_packed(make_tuple(N, C0, Hi, Wi, C1 / InWeiVectorSize));
+        make_naive_tensor_descriptor_packed(make_tuple(N, C0, Hi, Wi, C1));
     const auto wei_k_c0_y_x_c1_desc =
-        make_naive_tensor_descriptor_packed(make_tuple(K, C0, Y, X, C1 / InWeiVectorSize));
+        make_naive_tensor_descriptor_packed(make_tuple(K, C0, Y, X, C1));
     const auto out_n_k0_ho_wo_k1_desc =
         make_naive_tensor_descriptor_packed(make_tuple(N, K0, Ho, Wo, K1));
 
+    static_assert(in_n_c0_hi_wi_c1_desc.IsKnownAtCompileTime(), "");
+    static_assert(wei_k_c0_y_x_c1_desc.IsKnownAtCompileTime(), "");
+    static_assert(out_n_k0_ho_wo_k1_desc.IsKnownAtCompileTime(), "");
+
     constexpr auto conv_driver =
         DriverDynamicConvolutionForwardImplicitGemmDlops_v5r1_nc0hwc1_kc0yxc1_nk0hwk1_outpad<
             BlockSize,

host/driver_offline/include/driver_convolution_forward_implicit_gemm_v5r1_dlops_nc0hwc1_kc0yxc1_nk0hwk1.hpp

@@ -61,7 +61,7 @@ struct DriverDynamicConvolutionForwardImplicitGemmDlops_v5r1_nc0hwc1_kc0yxc1_nk0
         const auto C0 = in_n_c0_hi_wi_c1_global_desc.GetLength(I1);
         const auto Hi = in_n_c0_hi_wi_c1_global_desc.GetLength(I2);
         const auto Wi = in_n_c0_hi_wi_c1_global_desc.GetLength(I3);
-        // const auto C1 = in_n_c0_hi_wi_c1_global_desc.GetLength(I4);
+        //const auto C1 = in_n_c0_hi_wi_c1_global_desc.GetLength(I4);
 
         const auto K0 = out_n_k0_ho_wo_k1_global_desc.GetLength(I1);
         const auto Ho = out_n_k0_ho_wo_k1_global_desc.GetLength(I2);
@@ -78,11 +78,13 @@ struct DriverDynamicConvolutionForwardImplicitGemmDlops_v5r1_nc0hwc1_kc0yxc1_nk0
         const auto ConvDilationH = conv_dilations[I0];
         const auto ConvDilationW = conv_dilations[I1];
 
-        // const auto Hop = Number<(Ho + HoPerBlock - 1) / HoPerBlock * HoPerBlock>{};
-        // const auto Wop = Number<(Wo + WoPerBlock - 1) / WoPerBlock * WoPerBlock>{};
-
+#if 1
+        const auto Hop = Number<(Ho + HoPerBlock - 1) / HoPerBlock * HoPerBlock>{};
+        const auto Wop = Number<(Wo + WoPerBlock - 1) / WoPerBlock * WoPerBlock>{};
+#else
         const auto Hop = (Ho + HoPerBlock - 1) / HoPerBlock * HoPerBlock;
         const auto Wop = (Wo + WoPerBlock - 1) / WoPerBlock * WoPerBlock;
+#endif
 
         const auto OutRightPadH = Hop - Ho;
         const auto OutRightPadW = Wop - Wo;
@@ -119,8 +121,8 @@ struct DriverDynamicConvolutionForwardImplicitGemmDlops_v5r1_nc0hwc1_kc0yxc1_nk0
                                         make_tuple(Sequence<0, 1>{}, Sequence<2>{}, Sequence<3>{}));
 
         // input tensor
-        const auto in_n_c0_hip_wip_c1_global_desc = transform_tensor_descriptor(
-            in_n_c0_hi_wi_c1_global_desc,
+        const auto in_n_c0_hip_wip_e2_global_desc = transform_tensor_descriptor(
+            make_naive_tensor_descriptor_packed(make_tuple(K, C0, Hi, Wi, E2)),
             make_tuple(make_pass_through_transform(N),
                        make_pass_through_transform(C0),
                        make_pad_transform(Hi, InLeftPadH, InRightPadH),
@@ -129,8 +131,8 @@ struct DriverDynamicConvolutionForwardImplicitGemmDlops_v5r1_nc0hwc1_kc0yxc1_nk0
             make_tuple(Sequence<0>{}, Sequence<1>{}, Sequence<2>{}, Sequence<3>{}, Sequence<4>{}),
             make_tuple(Sequence<0>{}, Sequence<1>{}, Sequence<2>{}, Sequence<3>{}, Sequence<4>{}));
 
-        const auto in_n_c0_y_ho_x_wo_c1_global_desc = transform_tensor_descriptor(
-            in_n_c0_hip_wip_c1_global_desc,
+        const auto in_n_c0_y_ho_x_wo_e2_global_desc = transform_tensor_descriptor(
+            in_n_c0_hip_wip_e2_global_desc,
             make_tuple(
                 make_pass_through_transform(N),
                 make_pass_through_transform(C0),
@@ -142,7 +144,7 @@ struct DriverDynamicConvolutionForwardImplicitGemmDlops_v5r1_nc0hwc1_kc0yxc1_nk0
                 Sequence<0>{}, Sequence<1>{}, Sequence<2, 3>{}, Sequence<4, 5>{}, Sequence<6>{}));
 
         const auto b_e_n_ho_wo_e2_grid_desc = transform_tensor_descriptor(
-            in_n_c0_y_ho_x_wo_c1_global_desc,
+            in_n_c0_y_ho_x_wo_e2_global_desc,
             make_tuple(make_merge_transform(make_tuple(C0, Y, X)),
                        make_pass_through_transform(N),
                        make_pass_through_transform(Hop),
@@ -251,9 +253,9 @@ struct DriverDynamicConvolutionForwardImplicitGemmDlops_v5r1_nc0hwc1_kc0yxc1_nk0
                                   Sequence<0, 0, 0, 0, 0, 0, 0, 0, 0>{}));
         // clang-format on
 
-        // static_assert(a_e0_e1_k_e2_grid_desc.IsKnownAtCompileTime(), "");
-        // static_assert(b_e0_e1_n_ho_wo_e2_grid_desc.IsKnownAtCompileTime(), "");
-        // static_assert(c_k_n_hop_wop_grid_desc.IsKnownAtCompileTime(), "");
+        static_assert(a_e0_e1_k_e2_grid_desc.IsKnownAtCompileTime(), "");
+        static_assert(b_e0_e1_n_ho_wo_e2_grid_desc.IsKnownAtCompileTime(), "");
+        static_assert(c_k_n_hop_wop_grid_desc.IsKnownAtCompileTime(), "");
 
         // GEMM
         using GridwiseGemm = GridwiseGemmDlops_km_kn_mn_v3<

host/driver_offline/src/conv_fwd_driver_offline_nchwc.cpp

@@ -15,7 +15,7 @@
 #include "device_tensor.hpp"
 #include "device_convolution_forward_implicit_gemm_v5r1_dlops_nc0hwc1_kc0yxc1_nk0hwk1.hpp"
 
-#define USE_DYNAMIC_MODE 1
+#define USE_DYNAMIC_MODE 0
 #define USE_CONV_FWD_V5R1_NCHWC 1
 
 enum ConvForwardAlgo
@@ -90,7 +90,7 @@ int main(int argc, char* argv[])
     const bool do_log          = std::stoi(argv[4]);
     const int nrepeat          = std::stoi(argv[5]);
 
-#if 1
+#if 0
     constexpr auto N           = Number<1>{};
     constexpr auto C0          = Number<2>{};
     constexpr auto Hi          = Number<1080>{};
@@ -100,14 +100,16 @@ int main(int argc, char* argv[])
     constexpr auto Y           = Number<3>{};
     constexpr auto X           = Number<3>{};
     constexpr auto K1          = Number<8>{};
-#elif 0
+#elif 1
     constexpr auto N  = Number<1>{};
-    constexpr auto C  = Number<16>{};
+    constexpr auto C0 = Number<2>{};
     constexpr auto Hi = Number<540>{};
     constexpr auto Wi = Number<960>{};
-    constexpr auto K  = Number<64>{};
+    constexpr auto C1 = Number<8>{};
+    constexpr auto K0 = Number<8>{};
     constexpr auto Y  = Number<3>{};
     constexpr auto X  = Number<3>{};
+    constexpr auto K1 = Number<8>{};
 #elif 0
     constexpr auto N  = Number<1>{};
     constexpr auto C  = Number<16>{};