Personal computers (PCs) installed with specific system-on-a-chip (SoC) capabilities designed to run generative artificial intelligence (AI) tasks locally are expected to account for 60% of all PC shipments by the end of 2027.

This is according to a new forecast from International Data Corporation (IDC), which indicates that shipments of these AI PCs will increase from nearly 50-million units in 2024 to more than 167-million in 2027.

“As we enter a new year, the hype around generative AI has reached a fever pitch, and the PC industry is running fast to capitalise on the expected benefits of bringing AI capabilities down from the cloud to the client,” says IDC devices and consumer research group VP Tom Mainelli.

The ‘Worldwide artificial intelligence PC forecast 2023 to 2027’, IDC's first forecast for the worldwide AI PC market, predicts that, during 2024, AI PC shipments will ramp up, and over the next few years the technology is expected to move from niche to a majority.

The strong information technology decision-maker interest in AI PCs is being driven by the potential enhanced user productivity through faster performance, along with lower inferencing costs, and the benefit of on-device privacy and security.

“There are three primary technical reasons for bringing AI workloads from the cloud to the client: to enhance performance by eliminating the round trip that current AI workloads must make to the cloud and back over the network; to enhance privacy and security by keeping data on the device versus in motion; and to lower the cost by limiting the need to access costly cloud resources.”

Running an AI task locally on a PC, until recently, was done on the central processing unit, the graphics processing unit (GPU), or a combination of both, which can have a negative impact on the PC’s performance and battery life as the chips are not optimised to run AI efficiently.

“PC silicon vendors have now introduced AI-specific silicon to their SoCs called neural processing units (NPUs) that run these tasks more efficiently,” says Mainelli.

The IDC has, to date, identified three types of NPU-enabled AI PCs, namely hardware- enabled AI PCs, next-generation AI PCs and advanced AI PCs.

According to the IDC, hardware-enabled AI PCs include an NPU that offers less than 40 tera operations per second performance and typically enables specific AI features within apps to run locally.