https://github.com/hollance/bnns-vs-mpscnn
Compares the speed of Apple's two deep learning frameworks: BNNS and Metal Performance Shaders
https://github.com/hollance/bnns-vs-mpscnn
accelerate deep-learning ios metal neural-network
Last synced: 7 months ago
JSON representation
Compares the speed of Apple's two deep learning frameworks: BNNS and Metal Performance Shaders
- Host: GitHub
- URL: https://github.com/hollance/bnns-vs-mpscnn
- Owner: hollance
- Created: 2017-02-07T15:26:34.000Z (over 8 years ago)
- Default Branch: master
- Last Pushed: 2017-02-07T15:28:18.000Z (over 8 years ago)
- Last Synced: 2024-06-07T08:32:27.726Z (over 1 year ago)
- Topics: accelerate, deep-learning, ios, metal, neural-network
- Language: Swift
- Size: 118 KB
- Stars: 60
- Watchers: 8
- Forks: 11
- Open Issues: 0
-
Metadata Files:
- Readme: README.markdown
Awesome Lists containing this project
README
# BNNS vs Metal CNN benchmark
This app compares the speed of Apple's two deep learning frameworks: BNNS and Metal Performance Shaders (MPSCNN).
It creates a basic convolutional neural network with 2 convolutional layers, 2 pooling layers, and a fully-connected layer. Then it measures how long it takes to sends the same image 100 times through the network.
To run the app you need Xcode 8 and an iOS 10-compatible device with at least an A8 processor.
See also the [blog post](http://machinethink.net/blog/apple-deep-learning-bnns-versus-metal-cnn/).