For more than a year now, we have all been getting to grips with news about generative AI services that have emerged since the launch of ChatGPT. One of the issues with this technology right now is that all the processing is done in the cloud, but device manufacturers are planning to change that with local processing of AI content.
According to the analyst firm IDC, AI PCs and generative AI (GenAI) smartphones will total 295 million units by the end of this year, up from just 29 million units in 2023. These devices are equipped with dedicated AI accelerators or cores, neural processing units (NPUs), and accelerated processing units (APUs) or tensor processing units (TPUs) that have been designed to process AI tasks on-device.
Neowin has previously reported that PC vendors are looking to launch AI PCs around June and that Microsoft could be focusing its upcoming Windows 11 24H2 update on AI PCs.
Companies behind large language models often create smaller LLMs that are designed to run locally on devices, a popular example is Meta’s Llama 2 7B which is smaller than the 13B and 70B versions that are trained on more data.
“The rapid adoption of on-device GenAI capabilities and AI processors will eventually become a standard requirement for technology vendors,” said Ranjit Atwal, Senior Director Analyst at Gartner. “This ubiquity will pose challenges for vendors in differentiating themselves from competitors, making it harder to create unique selling points and drive increased revenues.”
GenAI smartphones will make up the bulk of shipments in 2024, with 240 million units. AI PC shipments are estimated to reach 54.5 million by the end of the year. Despite this difference, both will make up 22% of the market share of shipments in their respective markets.
In the case of AI PCs, IDC doesn’t expect this to drive end-user spending “beyond anticipated price increases.” It also said that software vendors will need time to utilize on-device AI and show its benefits to users.
Source: IDC