No
Yes
View More
View Less
Working...
Close
OK
Cancel
Confirm
System Message
Delete
Schedule
An unknown error has occurred and your request could not be completed. Please contact support.
Scheduled
Wait Listed
Personal Calendar
Speaking
Conference Event
Meeting
Interest
Schedule TBD
Conflict Found
This session is already scheduled at another time. Would you like to...
Loading...
Please enter a maximum of {0} characters.
Please enter a maximum of {0} words.
must be 50 characters or less.
must be 40 characters or less.
Session Summary
We were unable to load the map image.
This has not yet been assigned to a map.
Search Catalog
Reply
Replies ()
Search
New Post
Microblog
Microblog Thread
Post Reply
Post
Your session timed out.
This web page is not optimized for viewing on a mobile device. Visit this site in a desktop browser to access the full set of features.
2017 GTC San Jose

S7733 - Use NPP to Accelerate the Training Process

Session Speakers
Session Description

Learn how to use NVIDIA Performance Primitives (NPP) to leverage the GPU's power and accelerate training processes significantly. NPP is composed of over 5,000 signal and image processing functions, including GPU-based JPEG encoder and decoder functions, resize, warp affine transform, rotation, color conversion, etc. These functions can be used seamlessly in many areas of deep learning, such as image loading, data augmentation, etc. Not only these GPU functions are faster than their CPU counterparts, but also by using them the end users save the overhead from exchanging the memory between host and device. We'll introduce NPP and its performance against the state-of-the-art CPU implementations. Then we'll focus on showing how easy and natural it is to integrate NPP in deep training projects, with only minor changes to code bases. Finally, we'll take an example and present the performance comparison with and without NPP acceleration.


Additional Session Information
All
Talk
Deep Learning and AI Video and Image Processing
Automotive Cloud Services Software
25 minutes
Session Schedule