Bloomberg claims Apple wants AI hardware to match Google and Qualcomm
Apple is developing the chips for augmented reality and self-driving cars
Apple devices currently split AI tasks between the main processor and a GPU
New chip, known as the Apple Neural Engine, has its own module for AI requests
By Phoebe Weston For Mailonline
Apple will install a chip for carrying out artificial intelligence processing on future iPhones, according to a new report.
The chip, called the Apple Neural Engine, will bring new capabilities to mobile devices that will revolutionise how we use our phones, inside sources suggest.
The AI-enabled processor will handle tasks normally only managed by humans such as facial and speech recognition and could even alert us if it thinks we are getting ill.
The new chip, called the Apple Neural Engine, would bring new capabilities to mobile devices that will revolutionise how we use our phones, inside sources suggest. Pictured are leaked images of what the iPhone 8 might look like
WHAT COULD THIS MEAN?
The chip will handle tasks are normally only managed by humans such as facial and speech recognition.
An AI-enabled processor would help Apple get more advanced capabilities into their devices such as cars that self-drive and gadgets that have augmented reality.
In the future they could be used to monitor body readings and alert our phone when if we have an abnormal reading that needs acting on.
For example they could pick up early signs of conditions such as dementia, Parkinson's or diabetes.
This could be ascertained from limitless data on our every move - such as how we walk or talk.
The introduction of chips into phones will greatly speed up how fast machines can learn which will make AI more sophisticated than ever.
Apple engineers are trying to keep up with Google and Qualcomm, according to the report by Bloomberg.
Although Apple has got voice recognition assistant Siri, Amazon Echo and Google Home assistants have given other companies an advantage.
An AI-enabled processor would help Apple get more advanced capabilities into their devices such as cars that self-drive and gadgets that have augmented reality.
'Two of the areas that Apple is betting its future on require AI,' said Gene Munster, former Apple analyst.
'At the core of augmented reality and self-driving cars is artificial intelligence.'
Although AI is used to power assistants such as Siri these rely on computer servers processing data rather than processing data themselves.
The inclusion of AI software in phones will further increase how much we depend on them.
For example they could pick up early signs of conditions such as dementia, Parkinson's or diabetes.
This could be ascertained from limitless data on our every move - such as how we walk or talk.
Big Brother-style satellites capable of tracking CARS from...
The iPhone killer? Android inventor Andy Rubin unveils the...
Our best look yet at the iPhone 8? Leaks suggest Apple's new...
Currently Apple devices have two chips - a main processor and a graphics chip.
The new chip would mean these tasks would be carried out a single chip which would work even when the device is not connected to the internet.
The introduction of chips into phones will greatly speed up how fast machines can learn which will make AI more sophisticated than ever.
Google's hardware, the Tensor Processing Unit, is 15 to 30 times faster than the processors that power today's computers and is the company's first chip, announced in 2016.
Google's hardware, the Tensor Processing Unit (pictured), is 15 to 30 times faster than the processes that power today's computers
This unit connects Google services such as Google Search, Google Street View, Google Photos and Google Translate.
WHAT IS A TPU?
Google's hardware, the Tensor Processing Unit, is 15 to 30 times faster than the processors that power today's computers and is the company's first chip, announced in 2016.
TPUs are application-specific integrated circuits (ASICs) developed specifically for machine learning.
TPUs are designed for a higher volume of reduced precision computation than other types of processing units.
The chip was specifically designed for Google's TensorFlow framework, but Google still uses other chips for different types of machine learning.
At the I/O conference this year Google announced an updated version would soon be available to users of its cloud business and sources suggest it will soon be integrated into many devices.
AI on a phone is also likely to be useful in helping to keep the device secure and protected from external threats.
They could be used as a way to uniquely identify the wearer of a device.
Phones could be part of a broader network of machines which are dependent on AI - including cars.
Last month the UK government handed £12.8 million ($16.4 million) to an AI firm to develop a driverless car system.
The report suggests Apple may choose to discuss its latest AI work in its developers conference in June.
No comments:
Post a Comment