The Mobile Phone becomes smarter to perform several tasks autonomously without the intervention of the user, this is called deep learning or machine learning or Artificial Intelligence AI.
All About NPU? : ( Neural Processing Unit)
The NPU or Neural Processing Unit it is part of the SoC System on Chip or Mobile Processor these the first responsible of any smart action in your Phone it's simply the brains of your Smartphone, Welcom to Artificial Intelligence AI.
Historical:
in the past is delivered remotely only with servers Cloud, for example when you put a photo in the gooole translator it captures the text in it then do the translation, it's Google's AI that does this operation by using their cloud servers remotely, the same whene you use Google Lens for image recognition, now it's integrated directly into the Chip (SoC) Mobile Processor.
How the NPU Work?:
Inspired by how the human brain works, an NPU will learn about you and the way you use your smartphone, then perform intelligent actions such as shutting down unused apps to save power, suggesting text when you write, facial recognition...etc providing the user with new levels of smart vision, energy saving, security and more
for example Bixby Vision in Samsung Galaxy S20, S20 +, S20 Ultra and S21 series, detects the objects that you see in the camera!
and for the front camera you find also the Bokeh Effects (blur the background of your selfies) , It's the NPU that makes all these magic effects.
Some Tasks that the NPU Processor Can Do:
- Camera AI : NPU helps identify whether you are taking a panorama or a person and automatically adjusts the type of filtering used to give you the best results.
- Voice Recognition : it can identify user's voice to perform some tasks.
- Searching by Images.
- Facial recognition for security.
- Learning about your daily use of your phone to perform certain tasks to save Battery...etc.
Characteristics and operation of the NPU:
- NPU composed of multiple Processing Elements (PEs) that needed to run machine learning (control and arithmetic logic), with parallel architecture like GPU .
- NPU can compute machine learning tasks 10,000 times compared to GPUs (for Example the Huawei Kirin 980 processor used in the P30 Pro smartphone have a dual NPU core, capable of recognizing 4500 images per minute!) .
- As GPUs have APIs (out-of-the-box functions) that will be processed by the GPU (OpenGL, DirectX), NPUs have Frameworks (a group of libraries written by Python or C++ language that offers APIs for facilitates Machine Learning and create artificial intelligence applications) the most popular :Caffe,Caffe2, TensorFlow ,TensorflowLite,ONNX, Core ML(Apple).
3 Comments
I will share it with my other friends as the information is really very useful. Keep sharing your excellent work. Read more info about Runescape Private Server
ReplyDeleteThank you very much dear visitor, welcome to our blog I hope that I share with you useful information with a simplified method, Best regards Ramzy 😊
DeleteThis post is really awesome. Genuinely i like this blog. It gives me more useful information. I hope you share lots of things with us .RSPS instant PK
ReplyDelete