Additional initializations for testing

test/mx_mfma_op/mx_mfma_op.hpp

@@ -449,8 +449,33 @@ struct TestMFMA
         Tensor<CDataType> c_m_n_device_result(
             f_host_tensor_descriptor(params.M, params.N, params.StrideC, CLayout{}));
 
-        a_m_k.GenerateTensorValue(GeneratorTensor_1<ADataType>{0.015625f});
-        b_n_k.GenerateTensorValue(GeneratorTensor_Sequential<BDataType, 1>{});
+        switch(0)
+        {
+        case 0:
+            a_m_k.GenerateTensorValue(GeneratorTensor_1<ADataType>{0.015625f});
+            // NOTE: not all numbers are representable in FP8, BF8, etc.
+            b_n_k.GenerateTensorValue(GeneratorTensor_Sequential<BDataType, 1>{});
+            break;
+        case 1:
+            a_m_k.GenerateTensorValue(GeneratorTensor_1<ADataType>{1.0f});
+            b_n_k.GenerateTensorValue(GeneratorTensor_1<BDataType>{1.0f});
+            break;
+        case 2:
+            // expect small round off errors
+            a_m_k.GenerateTensorValue(GeneratorTensor_3<ADataType>{-5, 5});
+            b_n_k.GenerateTensorValue(GeneratorTensor_3<BDataType>{-5, 5});
+            break;
+        case 3:
+            // expect small round off errors
+            a_m_k.GenerateTensorValue(GeneratorTensor_4<ADataType>(-1, 3));
+            b_n_k.GenerateTensorValue(GeneratorTensor_4<BDataType>(1, 3));
+            break;
+        default:
+            a_m_k.GenerateTensorValue(GeneratorTensor_2<ADataType>{-5, 6});
+            b_n_k.GenerateTensorValue(GeneratorTensor_2<BDataType>{-5, 6});
+
+            break;
+        }
 
         return std::make_tuple(a_m_k, b_n_k, c_m_n_host_result, c_m_n_device_result);
     }