Intel sees artificial intelligence PCs reaching half of shipments by 2026

Intel expects artificial intelligence PCs, defined by the presence of dedicated neural processing hardware, to approach half of global PC shipments by 2026, even as buyers prioritize performance and battery life over artificial intelligence features.

Intel’s Japan president Makoto Ohno is positioning the current year as a decisive moment for the adoption of artificial intelligence PCs, forecasting a rapid expansion of systems equipped with dedicated neural processing hardware. He indicated that artificial intelligence PCs are set to grow from a niche category into a mainstream segment within the broader PC market in a relatively short timeframe. The comments highlight both optimism about the technology’s trajectory and realism about what is driving consumer demand today.

Ohno stated that by 2026, artificial intelligence PCs will account for roughly half of the unit shipments for the year. He explained that out of the total unit shipments, which IDC estimates at around 260 million, about 130 million are expected to be artificial intelligence PCs. The term is described as loosely defined, referring to PCs that come with an NPU or other artificial intelligence processing unit, which allows the system to perform some local artificial intelligence processing with enough power to meet basic user needs. This hardware focus underscores an industry push to embed on-device artificial intelligence capabilities as a standard feature in new systems.

At the same time, Ohno emphasized that current purchase decisions are driven less by interest in artificial intelligence functions and more by core computing benefits delivered by the latest chips. He noted that the primary reason for purchasing an artificial intelligence PC may not be its artificial intelligence capabilities, but rather the better battery life due to newer and more optimized silicon. He added that it is also predicted that by 2026, this will exceed half, with one in two computers being an artificial intelligence PC, yet reasons for choosing such a system include its high performance and the long battery life provided by the use of an NPU. In other words, people are not currently purchasing an artificial intelligence PC in order to use its artificial intelligence related functions, even as the installed base of capable machines grows.

55

Impact Score

Nvidia Blackwell Ultra targets agentic artificial intelligence with higher efficiency

Nvidia’s Blackwell Ultra platform is being positioned to power a surge in agentic artificial intelligence and coding assistants, with new data indicating major gains in throughput and cost efficiency over Hopper. Inference providers are adopting the architecture to handle rising software development workloads that demand low latency and long context processing.

The power struggles behind predictive technology

Three new books trace how modern predictive technologies, from supervised learning algorithms to rational choice theory, have turned forecasts into tools of power and control. Together they argue that democratic oversight, human judgment, and resistance to self-fulfilling prophecies are essential correctives to an increasingly automated future.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.