@@ -1,2 +1,5 @@
https://www.tensorflow.org/guide/create_op
+https://www.elektronikpraxis.vogel.de/welche-hardware-eignet-sich-besser-zur-ki-beschleunigung-a-720869/
+https://www.extremetech.com/computing/247199-googles-dedicated-tensorflow-processor-tpu-makes-hash-intel-nvidia-inference-workloads
+https://towardsdatascience.com/types-of-optimization-algorithms-used-in-neural-networks-and-ways-to-optimize-gradient-95ae5d39529f
