Get our free extension to see links to code for papers anywhere online!

 Add to Chrome

 Add to Firefox

CatalyzeX Code Finder - Browser extension linking code for ML papers across the web! | Product Hunt Embed

Optimizing Memory Efficiency for Deep Convolutional Neural Networks on GPUs

Oct 12, 2016
Chao Li, Yi Yang, Min Feng, Srimat Chakradhar, Huiyang Zhou

Leveraging large data sets, deep Convolutional Neural Networks (CNNs) achieve state-of-the-art recognition accuracy. Due to the substantial compute and memory operations, however, they require significant execution time. The massive parallel computing capability of GPUs make them as one of the ideal platforms to accelerate CNNs and a number of GPU-based CNN libraries have been developed. While existing works mainly focus on the computational efficiency of CNNs, the memory efficiency of CNNs have been largely overlooked. Yet CNNs have intricate data structures and their memory behavior can have significant impact on the performance. In this work, we study the memory efficiency of various CNN layers and reveal the performance implication from both data layouts and memory access patterns. Experiments show the universal effect of our proposed optimizations on both single layers and various networks, with up to 27.9x for a single layer and up to 5.6x on the whole networks.

* Published as a conference paper International Conference on High Performance Computing, Networking, Storage, and Analysis (SC'16), 2016 

Share this with someone who'll enjoy it:

   Access Paper Source

Share this with someone who'll enjoy it: