Optimized Fc Impl

include/aidge/backend/cpu/operator/FCImpl_kernels.hpp

@@ -96,21 +96,16 @@ void FCImpl_cpu_forward_kernel(const DimSize_t batchSize,
     const B* biases = static_cast<const B*>(biases_);
     O* output = static_cast<O*>(output_);
 
-    if (biases == nullptr) {
-        std::fill(output, output+(batchSize*outputFeatureSize), B(0));
-    }
-    else {
-        for (std::size_t batch = 0; batch < batchSize; ++batch) {
-            std::copy(biases, biases+outputFeatureSize, output+(batch*outputFeatureSize));
-        }
-    }
-
-    for (std::size_t batch = 0; batch < batchSize; ++batch) {
-        for (std::size_t out = 0; out < outputFeatureSize; ++out) {
+#ifdef _OPENMP
+    #pragma omp parallel for collapse(2) if (batchSize * outputFeatureSize > 32)
+#endif
+    for (int batch = 0; batch < static_cast<int>(batchSize); ++batch) {
+        for (int out = 0; out < static_cast<int>(outputFeatureSize); ++out) {
+            const auto biasVal = (biases) ? biases[out] : B(0);
             output[out + batch*outputFeatureSize] = std::inner_product(input + batch*inputFeatureSize,
                                                         input + (batch + 1)*inputFeatureSize,
                                                         weights + out*inputFeatureSize,
-                                                        output[out + batch*outputFeatureSize]);
+                                                        biasVal);
         }
     }
 }