Some of us use things, but we don’t care to know the origin or how it came to be. Anyway, a topic that will teach you how mobile phones evolved from feature phones to smartphones will only interest a nerd or an individual that’s preparing for an interview scheduled at Samsung or any other mobile phone company.
The company that spearheaded the evolution of mobile phones; evolving from feature phones to smartphones now want to push their effort a little bit further and make smartphones smarter.
This is a story of a company that’s about to give wings to a creation it created to walk. I am referring to “ARM” which stands for “Acorn RISC Machine”. The British chip design firm that came up with the design for the processors that’s currently used in almost all the world’s smartphones today. When you see any smartphone that its design wasn’t modelled from their original chip design, know that that smartphone is not a smartphone.
The chip maker company, ARM is planning to add a particular type of hardware into their new mobile phone chip design that will let your smartphone run artificial-intelligence algorithms.
A few days ago, ARM announced that it created the world’s first dedicated machine-learning chips for use in smartphones and smart-home devices.
According to our source, the company will be sharing the plans or blueprint with its hardware partner, smartphone chipmaker Qualcomm, and expects to see devices being packaged with the hardware show up in the smartphone market by early 2019.
Currently, we have smartphones that use machine learning while implementing certain A.I features. These devices lack the horsepower to run A.I algorithms, so they work by using the help of big servers located in the cloud where codes are stored and utilized from.
Making chips that enable mobile devices to run their own AI software is not only attractive, but it will also speed up things and cut the lag that will come when smartphone chips send and receive information back and forth from cloud servers. It will also allow hardware to run offline, bringing confidence to individuals that advocate too much about privacy. Who wouldn’t be comforted when his/her mobile phone’s data remains on the device?
According to Jem Davies, the Engineer who leads the machine-learning group at ARM, he acknowledged in an interview that the company has spent a long time and also resources in getting the chips to run AI software efficiently.
He stated that the new chip uses less power than any other chip designed by ARM to perform the linear-algebra calculations that modern artificial intelligence programs are known for. Also, they are very good at moving data in and out of memory.
It is worth knowing that ARM isn’t the only company trying to innovate on mobile AI chips. The iPhone X that we all know contains a “neural engine” as part of its main chipset, which Apple created to run artificial neural networks for things like images and speech processing. So, Apple is also in this game.
Huawei’s Mate 10 smartphone contains a similar in-house built chip that is called a neural processing unit. So, Huawei is also here.
Google’s Pixel 2 smartphone also has a chipset built by Google itself to help tackle imaging and machine-learning problems.
Even with the impending competition ARM’s new chip is going to encounter when it’s eventually ready, ARM has kept an impressive record of designing energy-efficient processors for mobile applications more than any other company in the world. Therefore, smartphone manufacturers are bound to adopt this new chip in their devices ones again. We should all expect ARM’s new A.I brains in plenty of devices next year.