terminology clean-up (#1792)

include/ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp

@@ -1544,7 +1544,7 @@ struct ThreadwiseTensorSliceTransfer_StaticToStatic
     ElementwiseOperation element_op_;
 };
 
-// Specilized for WMMA-Navi3
+// Specialized for gfx11
 // A single Wave32 is composed by double row
 // Data exchange allowed between these two rows
 // This RowLane Dst buf will be filled from two Src buf
@@ -1679,7 +1679,7 @@ struct ThreadwiseTensorSliceTransfer_StaticToStatic_InterRow
     ElementwiseOperation element_op_{};
 };
 
-// Specilized for WMMA-Navi4
+// Specialized for gfx12
 template <typename SrcData,
           typename DstData,
           typename SrcDesc,

include/ck/tensor_operation/gpu/warp/wmma_gemm.hpp

@@ -307,7 +307,7 @@ struct wmma_type<WmmaInstr::wmma_f32_16x16x16_f16_gfx12,
 
     // Wave mode dependent propety
     static constexpr index_t wave_size = Number<WaveSize>{};
-    // * Fixed in Navi3x, Will be wave mode dependent on Navi4x
+    // * Fixed for gfx11, Will be wave mode dependent on gfx12
     // static constexpr index_t num_src_a_vgprs_per_wave = k_per_wmma / 2 * src_a_data_size / 4;
     // static constexpr index_t num_src_b_vgprs_per_wave = k_per_wmma / 2 * src_b_data_size / 4;
     // * num_acc_vgprs_per_wave alone M direction