SOURCE RESEARCH
Artificial Intelligence and Machine Learning: Personal & Work Devices
Market Data | 3Q 2025 | MD-AIMLED-103
Get The ReportHeterogeneous AI chipsets distributing AI workloads between the CPU, GPU, and NPU (or equivalent AI accelerator) will dominate overall shipments as we move to the end of the decade, as these have emerged, and are proving themselves to be, the most efficient systems for addressing both traditional and novel Gen AI workloads. This is evidenced by the growing Independent Software Vendor (ISV) ecosystems like Intel’s AI PC Acceleration Program and Qualcomm’s AI Hub.
It remains important that the ISV ecosystem aligns with the trajectory by chipset vendors, Original Equipment Manufacturers (OEMs) and Operating System (OS) platforms (Windows, iOS, Android) to develop applications that take full advantage of on-device AI hardware and deliver productivity AI and feature-enhancing applications. Device sales will only continue to enjoy an uplift if consumers and enterprise customers see the advantage of on-device AI being leveraged to limit networking costs, decrease latency, and retain users’ private data on the device.
