AlexNet made a big change to deep learning. A deep convolutional neural network (CNN) was used to handle the pictures. It had eight layers, with three fully linked layers and five convolutional layers. For regularization, AlexNet also used dropout and ReLU activation functions. To speed up training, it used GPU computing. It won a top competition in 2012 because it was so much better at recognizing images than older ways. This achievement demonstrated the usefulness of deep learning and sparked the creation of many more advanced networks and software used in picture recognition and other areas.