The impact of the AC922 Architecture on Performance of Deep Neural Network Training

The paper presents assessment of Unified Memory performance with data prefetching and memory oversubscription. Several versions of code are used with: standard memory management, standard Unified Memory and optimized Unified Memory with programmer-assisted data prefetching. Evaluation of execution times is provided for four applications: Sobel and image rotation filters, stream image processing and computational fluid dynamic simulation, performed on Pascal and Volta architecture GPUs-NVIDIA GTX 1080 and NVIDIA V100 cards. Furthermore, we evaluate the possibility of allocating more memory than available on GPUs and assess performance of codes using the three aforementioned implementations, including memory oversubscription available in CUDA. Results serve as recommendations and hints for other similar codes regarding expected performance on modern and already widely available GPUs.

show abstract

The impact of the AC922 Architecture on Performance of Deep Neural Network Training

Cited by 3 publications

References 8 publications

Food Classification from Images Using a Neural Network Based Approach with NVIDIA Volta and Pascal GPUs

Food Classification from Images Using a Neural Network Based Approach with NVIDIA Volta and Pascal GPUs

Adaptation of AI-Accelerated CFD Simulations to the IPU Platform

Performance evaluation of Unified Memory with prefetching and oversubscription for selected parallel CUDA applications on NVIDIA Pascal and Volta GPUs

Contact Info

Product

Resources

About