Cristiano Amon, a long-time veteran of Qualcomm chip giant and president of the company, is convinced that 5G wireless technology will allow you to see amazing things: a cellular network upgrade deployed by AT & T and others that still remains a mystery to the average consumer. "Let's go for a walk in the past," he proposed, in a conversation with ZDNet on the gigantic Qualcomm booth at the Consumer Electronics Show in Las Vegas this week.
Also: The best of CES 2019: a cool technology to buy this year
"Remember that at the beginning of 4G, everyone was saying: Why do I need a hundred megabit device per second?" he thinks. "The operators told people that it would be to connect a laptop.
"Now we're looking at our smartphones and we're saying, how could we live without that? You have to design a technology that will last 10 years."
The president of Qualcomm, Cristiano Amon.
"By offering developers a very low latency, it's a part of the picture," he says, referring to the average time required for the first bit of data to roam a cellular network, to make it easier for users to access the data. one phone to the other.
Also: CES 2019: Huang, CEO of Nvidia, explains how artificial intelligence changes everything
"You will get a phone – and there will be about thirty advertised models that will start from the second quarter – that will be at least ten times faster than today, with instant response time. What you take for granted Today, just like storing music in the cloud, video will be as easy to send and consume as music today on a smartphone: sports and news. that you want will be instantly accessible.
Also: How Facebook handles the fact that AI is a mess on smartphones
"All social media companies are very excited about this, you can broadcast messages to your friends, instead of checking their tweets, you will have an instant presence with your friends."
The second thing that comes with the 5G is a much more sophisticated partitioning of computer tasks, including artificial intelligence, he is convinced. While today, a little machine learning is downloaded on a device, in the future, says Amon, servers and cloud computing phones will work in tandem to process neural network models. "This server, with a low latency response, will become a co-processor for your phone, and consider 5G as a network bus connecting your phone to the cloud."
He extends to other areas beyond the AI, he says. "Imagine that your phone has the capabilities of a $ 10,000 gaming computer. You're rendering on the server, and then the result is sent to the phone screen." have on-demand capacity, depending on the workloads you need. "
Another area is that of storage: "All the decision made today about the capacity of my phone, its storage, will disappear as you have access to everything you want, in the cloud, on 5G. "
Also: The artificial intelligence on Android mobile phones remains a work in progress
In the fall, says Amon, it will be a little difficult not to buy a 5G phone. "You will go to an operator, you will say you want a new phone and you may want to buy a 4G phone, but you really have to work for it," he says, representing salespeople for the company. direct to a 5G device; they will do their best to convince you that 4G is not the solution, for the reasons mentioned above.
Towards the end of the conversation, Amon was asked about the current confusion of AI on mobile. As reported recently by researchers in artificial intelligence Facebook, the situation of chipsets is a mess. Not only is the market between Qualcomm and its competitors very fragmented, but the vast majority of chips running smartphone software in the wild are six years old or older. More sophisticated support for neural networks seems to be needed.
Amon nodded and said that machine learning would extend to many other parts of Qualcomm's portfolio. "The fastest growing silicon area for Snapdragon [Qualcomm’s mobile chipset brand] is the machine learning. And one thing we do is expand the capacity on several levels, as the use cases become more and more important. "
"You will see that this is affecting other parts of the smartphone market, as you can see, the capabilities are moving from one level to the other of the mobile."
"Over time, it will not be a problem anymore, we have the 800, 700 and 600 series. [of Snapdragon] Today, support machine learning with AI processing capabilities specifically. We can distribute the neural networks on the neural network block, sometimes on the processor and sometimes on the graphics processor. "
"And for the rest of the wallet, well, stay tuned."