The anticipated shipments of AI-equipped PCs, capable of running generative AI tasks locally, are poised to reach nearly 50 million units this year, with projections suggesting a significant threefold increase in just a few years.

Traditionally, AI tasks on local machines were managed by the CPU, GPU, or a combination of both, presenting challenges due to their suboptimal performance for AI workloads, impacting overall system efficiency and draining batteries quickly.

The emergence of Neural Processing Units (NPUs) designed explicitly for AI workloads is transforming this landscape. IDC categorizes NPU-enabled AI PCs into three classes:

  1. Hardware-enabled AI PCs: Equipped with NPUs offering under 40 tera operations per second (TOPS) performance, these systems typically enable AI features within applications. Leading players like Apple, AMD, Qualcomm, and Intel actively ship chips in this category.
  2. Next-generation AI PCs: Featuring NPUs with 40 – 60 TOPS performance, these PCs utilize AI-first operating systems, enabling system-wide AI capabilities. Announcements from AMD, Qualcomm, and Intel indicate the arrival of chips in this category later this year.
  3. Advanced AI PCs: Characterized by offering over 60 TOPS of performance, no vendors have currently unveiled products in this category. However, IDC anticipates their introduction in the coming years, although this forecast is not reflected in the latest IDC projections.

IDC foresees a rapid increase in shipments of hardware-enabled AI PCs over the next two years, paving the way for the dominance of next-gen AI PCs. By 2027, the research firm predicts that shipments of next-gen AI PCs will double those of hardware-enabled AI PCs. These cutting-edge systems are expected to cater to commercial clients, offering enhanced digital content creation tools and AI-infused PC gaming experiences, promising an exciting future for both business and consumer markets.

By Impact Lab