Today, major companies are working hard to change the underlying hardware of their products to quickly catch up with this trend. According to Bloomberg, Apple is actively producing sophisticated artificial intelligence processing chips to speed up the speed of artificial intelligence algorithms and use it to extend battery life.
Bloomberg reported that the chip, dubbed the "Apple Nervous System" within Apple, will be used for devices with facial and speech recognition tasks. The latest iPhone 7 equipped with artificial intelligence services is mainly related to the camera, some of which use an image signal processor and a graphics processing unit integrated on the A10 fusion chip. However, if you can use a more professional processor to run these algorithms more efficiently, Apple can not only speed up the speed of image recognition, but also save the power consumption of the iPhone battery. Bloomberg said Apple plans to integrate the chip into many devices and test the prototype of the chip on a future iPhone. Apple wants to test the chip in services such as photo applications, speech recognition and predictive keyboards.
Bloomberg said Apple will also provide developers with the opportunity to run a variety of artificial intelligence tasks on third-party iOS applications. This is not surprising. Many technology companies are trying to use a variety of hardware to optimize deep learning, because this is a major approach to artificial intelligence learning, which requires a large amount of data to learn and identify patterns. Currently, graphic expert Nvidia is an early leader in this regard. The GPU traditionally used to generate computer graphics is the company's main platform for training deep learning algorithms. Google has been deploying a custom chip called "tensor processing unit" in its back office, and will begin to provide services to external companies using Google's second-generation chip "cloud TPU" on Google Cloud. Currently, most artificial intelligence services are calculated in cloud computing. For example, voice speakers made by Amazon and Google need to connect to the Internet to transmit voice data to a huge cluster of computers to run deep learning algorithms on the data. However, more and more deep learning algorithms, also called inferences, need to start running on the device to reduce latency and improve privacy protection. Technology analyst Patrick Moorhead said: "Apple must make more inferences on the device, let the machine deep learning to protect people's privacy. By contrast, Google has arranged everything in the cloud."
As the leading chip maker for Android smartphones, Qualcomm has begun to optimize its Opteron processor in a variety of deep learning frameworks. Qualcomm runs these algorithms on a digital signal processor. It is widely believed that Apple's progress in artificial intelligence is lagging behind technology giants such as Google, Amazon and Microsoft. For example, when Siri was launched in 2011, Apple was the leader of voice artificial intelligence assistants, but now Google and Amazon's voice assistant technology is far ahead of Apple, they have more advanced technology to support Google Home and Amazon Echo.
It is rumored that Apple is developing its own new Siri intelligent voice service. But Apple did not respond to the reporter's request for comment.
Henan Yongrong Power Technology Co., Ltd , https://www.hnyongrongglobal.com