Strategy Analytics: Google leads in AI-enabled smartphones

A recent research report released by Strategy Analytics pointed out that one in three smartphones sold worldwide this year uses an artificial intelligence (AI)-based virtual assistant. However, cloud-based AI virtual assistants respond slowly - on-device machine learning capabilities will greatly improve the speed of smartphone AI.

Ville Ukonaho, senior analyst at Strategy Analytics, said, “In 2017, Google maintained a slight lead in the global market for smartphones with native virtual assistants. As Android smartphones with native Google Assistant continue to penetrate into lower price segments, Google’s lead will also expand.”

Figure 1 : Global market share of smartphones with virtual assistants by type : 2017 & 2022

Ville Ukonaho, senior analyst at Strategy Analytics, said, “In 2017, Google maintained a slight lead in the global market for smartphones with native virtual assistants. As Android smartphones with native Google Assistant continue to penetrate into lower price segments, Google’s lead will also expand.”

Over the past few years, since the first virtual assistant on smartphones, Siri, appeared on Apple iPhone, most of the data processing has been done in the cloud because of the limited processing power of smartphone CPUs. AI applications require a lot of data processing, even for small tasks. Only recently have smartphones started to have the computing power to process tasks locally. Upgrades to smartphone operating systems and related software and components have brought more powerful processing power to the latest flagship smartphones.

Ken Hyers, Research Director for Emerging Devices at Strategy Analytics, said: “Some vendors have created more advanced processing engines or combined the capabilities of CPUs, GPUs and DSPs to form subsystems capable of handling complex machine learning and other AI tasks.” However, only high-end flagship smartphones are equipped with these advanced AI processing engines and subsystems, which means that virtual assistants driven by local AI on the device will become a key differentiator for high-end smartphones.

“We expect the capabilities of virtual assistants to improve significantly over the next few years through a combination of software enhancements based on machine learning and hardware enhancements in the form of AI engines,” said Neil Mawston, Executive Director at Strategy Analytics. “This will result in faster responses from virtual assistants and more interactive experiences from the endpoint.”