Snapdragon 820 Deep Learning SDK Will Promise Better Privacy For The Future
2017-04-23 23:16
1026 查看
Snapdragon 820 deep learning SDK introduced by Qualcomm
0By Muhammad on
MAY 2, 2016Hardware
& Machines, Technology
FacebookTwitterGoogle+RedditPinterestEmail分享
Qualcomm announces new deep learning SDK with support for Snapdragon 820
With Snapdragon 820 being one of the most powerful chipsets that you see present in high-end Android smartphones, Qualcomm has decided to make the SoC even more than effective by launching a new software development kit called the Snapdragon Neural ProcessingEngine.
The company’s new SDK is built for Zeroth, which is a machine intelligence platform. According to additional details from The Verge, this software development kit will allow companies to run deep learning programs locally on your devices, with the most common
being your smartphone.
Not only this, but thanks to the new Snapdragon 820 SDK, you will be ensured better privacy as the system will not be uploading much data to the cloud, since all the deep learning will be done locally. However, for that to happen, your smartphone or tablet
must be equipped with a Snapdragon 820.
Some useful features of the Snapdragon Neural Processing Engine:
Accelerated runtime for on-device execution of convolutional and recurrent neural networks on the Snapdragon 820 cores (Qualcomm Kryo CPU, Qualcomm Adreno GPU, Qualcomm Hexagon DSP)Support for common deep learning model frameworks, including Caffe and CudaConvNet;
A lightweight, flexible platform designed to utilize Snapdragon heterogeneous cores to deliver optimal performance and power consumption;
Supports companies in a broad range of industries, including healthcare, automotive, security, and imaging, to run their own proprietary trained neural network models on portable devices.
Qualcomm’s SDK will not just be limited to smartphones and tablets. This is because companies can integrate it into any type of smart devices. The largest mobile chipset manufacturer also pointed out that Nauto, a startup that makes smart dashboard cameras
will also be integrating the new SDK into its devices, allowing the company to provide better security for drivers.
If the Snapdragon Neural Processing Engine is available for a Snapdragon 820, then it will also be available for Snapdragon 830, which is most likely going to be manufactured on the 10nm FinFET technology. Question is, will the new SDK support upcoming mid-ranged
chipsets? Looks like time will tell.
相关文章推荐
- How to Use the Keras Functional API for Deep Learning
- How can I know if Deep Learning works better for a specific problem than SVM or random forest?
- 实时SLAM的未来及与深度学习的比较The Future of Real-Time SLAM and “Deep Learning vs SLAM”
- 【论文阅读笔记】Deep Learning in Medical Imaging: Overview and Future Promise of an Exciting New Technique
- 泛读:CVPR2014:Discriminative Deep Metric Learning for Face Verification in theWild
- The Future of Real-Time SLAM and “Deep Learning vs SLAM”
- 翻译 | Keras : Deep Learning library for Tensorflow and Theano
- Discriminative Deep Metric Learning for Face Verification in the Wild(文献泛读)
- The Future of Real-Time SLAM and "Deep Learning vs SLAM"
- DEEP LEARNING IS THE FUTURE: Q&A WITH NAVEEN RAO OF NERVANA SYSTEMS
- The Future of Real-Time SLAM and Deep Learning vs SLAM
- An Introduction to Deep Learning for the PhysicalLayer
- How to install Theano on Amazon EC2 GPU instances for deep learning
- Some Problems for Deep Learning with Theano && Lasagne framework
- TensorLayer : The Fastest Deep Learning Library for TensorFlow
- The Promise of Deep Learning
- We're on the cusp of deep learning for the masses. You can thank Google later
- The Future of Real-Time SLAM and "Deep Learning vs SLAM" 深度学习与slam
- How to install the Nokia S60 SDK in NetBeans for J2ME development
- 解决"The breakpoint will not currently be hit. No symbols have been loaded for this document" 错误