Skip to main content [aditude-amp id="stickyleaderboard" targeting='{"env":"staging","page_type":"article","post_id":1670971,"post_type":"story","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"business,mobile,","session":"D"}']

Qualcomm shows off deep learning for mobile devices and ultrasonic fingerprint scanning

Qualcomm CEO Derek Aberle speaking during his company's press conference at MWC 2015.

Image Credit: Mark Sullivan/VB

BARCELONA, Spain — Qualcomm gave a broad overview of its key initiatives here at Mobile World Congress, but took special care to describe two new initiatives that focus on ultrasonic fingerprint identification and on-device deep learning.

Qualcomm’s new Snapdragon Sense ID 3D fingerprint technology is designed to enable easier and more secure authentication on all kinds of mobile devices.

[aditude-amp id="flyingcarpet" targeting='{"env":"staging","page_type":"article","post_id":1670971,"post_type":"story","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"business,mobile,","session":"D"}']

The new fingerprint reader uses ultrasonic readers to find unique contours and even sweat glands in the user’s finger, Qualcomm president Derek Aberle pointed out. He pointed out that ultrasonic is also more impervious to dirt and dust than existing fingerprint readers.

Qualcomm hopes this technology will eventually make passwords a thing of the past for authentication in mobile devices.

AI Weekly

The must-read newsletter for AI and Big Data industry written by Khari Johnson, Kyle Wiggers, and Seth Colaner.

Included with VentureBeat Insider and VentureBeat VIP memberships.

Deep learning on mobile

Qualcomm is also showing off its Zeroth Platform, a cognitive, or deep learning, chip technology for mobile devices. The technology allows for deep learning to happen natively on a mobile device, without having to call back for data in the cloud.

The new technology can be used in several ways, but Qualcomm focused on a couple of camera applications. For instance, the deep learning technology could learn to identify specific items in a photograph, and apply a set of ideal camera settings. The Zeroth technology can also be used to recognize people in photographs.

Zeroth also gets smarter over time as recognizes more things and people in photographs.

Above: The Zeroth technology learns to recognize specific features of photographs.

Image Credit: Mark Sullivan/VentureBeat

Both of these new mobile technologies will likely rely on Qualcomm’s newest line of processors — the Snapdragon 820. But Qualcomm said precious little about the new processor at the press conference here. The new processor is expected out in the second half of this year.

At the Qualcomm booth, the company is demonstrating its work on allowing mobile devices to hand over voice and video calls to Wi-Fi and LTE and vice versa. Qualcomm is working on aggregating Wi-Fi and LTE signals to enhance the connection speeds of mobile devices.

The company is also showing off its latest line of modems — Category 11 LTE modem chips that support 600 mbps download speeds.

[aditude-amp id="medium1" targeting='{"env":"staging","page_type":"article","post_id":1670971,"post_type":"story","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"business,mobile,","session":"D"}']

Above: Qualcomm’s new 820 processor is expected in the second half of this year.

Image Credit: Mark Sullivan/VentureBeat

Unlicensed spectrum

In a world where wireless spectrum is limited and precious, and where demand for data continues to spike up, chip makers and carriers are looking for new ways to relieve network congestion and maintain connection speeds.

One way to do that is by utilizing unlicensed spectrum for running LTE signal. “Your starting to see us supporting LTE-U (LTE in unused spectrum) in small scale SOC platforms,” Aberle said. Qualcomm has also announced a commercial RF transceiver for LTE-U.

Responding to claims that LTE in unlicensed spectrum interferes with other wireless signal, Aberle had this to say: “LTE is actually a better neighbor than another access point — there really is no interference issue at all.”

Above: HTC’s Peter Chou speaking at Qualcomm’s press conference in Barcelona Monday. HTC’s new One M9 runs on a Snapdragon 810 chip.

Image Credit: Mark Sullivan/VentureBeat

Fighting on all fronts

Qualcomm says it wants to be providing chips and modems for the automotive, mobile computing, smart cities, wearables, smart home, and networking industries.

[aditude-amp id="medium2" targeting='{"env":"staging","page_type":"article","post_id":1670971,"post_type":"story","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"business,mobile,","session":"D"}']

In automotive, Qualcomm has singed an agreement with Mercedes AMB Petrona Formula One in which it will provide the chips that power in-car sensor networks. The company now has connected car programs going with 15 automakers.

As for wearables, Qualcomm says its Snapdragon 400 chips now power most of the Android wearables on the market. Huawei just announced a new wearables initiative that will use Snapdragon 400 processors, Aberle said.

“We shipped 920 million chipsets in 2014,” Aberle said. “We are getting a lot of traction with the (Snapdragon) 810 processor.

HTC’s new One M9 phone runs on the 810, for example, and Aberle says another 60 phones are now being designed with the 810 built in.

[aditude-amp id="medium3" targeting='{"env":"staging","page_type":"article","post_id":1670971,"post_type":"story","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"business,mobile,","session":"D"}']

VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Learn More