G. Huang, Z. Liu, L. van der Maaten, and K. Q. Pytorch: An imperative style, high-performance deep learningĭ. Gyawali, A. Regmi, A. Shakya, A. Gautam, and S. Shrestha.Ĭomparative analysis of multiple deep cnn models for waste In 2009 21st International Symposium on Computer ArchitectureĪnd High Performance Computing, pages 11–18, 2009. Profiling general purpose gpu applications. Neural network model using tensorflow and a big dataset. ReferencesĬomparison between cpu and gpu for parallel implementation for a After profiling, we know how the performance is impacted by the use of CPU and GPU to handle the optimization procedure. From the experiment, we also conclude that some parameters of the neural network model are also responsible for resource consumption across CPU and GPU. We have studied the performance of different metrics regarding CPU and GPU usage. A lot of factors impact artificial neural network training. We have conducted experiments to show how the deep learning model CPU and GPU impact the time and memory consumption of CPU and GPU. ( 2015) describe the flexible software profiling of GPU architectures. ( 2009) mention the ways to optimize the GPU memory for large-scale datasets.
Salgado Salgado ( 2015) describes profiling kernel behavior to improve CPU/GPU performances. In the same way, Alkaabwi Alkaabwi ( 2021) uses Tensorflow and Big Dataset to compare CPU and GPU neural network parallel implementation. ( 2009) mention about the profiling strategy based on performance predicates and identify major causes of performance degradation.
They have used the Tensorflow profiler to trace the operations in CPU and GPU. ( 2020) talked about building deep CNN models across GPU for different algorithms. Lind ( 2019) talk about the performance comparison between CPU and GPU in Tensorflow and Gyawali et al. A number of works have been done to identify CPU and GPU performances over different algorithms and operations.