Intel and other major silicon vendors are targeting the enterprise market and developer space as the main opportunities for new AI technologies, according to a report from Omdia. The report, titled “AI PCs: An Emerging Device Class,” outlines two main technology options for delivering AI inference on PCs.
One option involves a traditional PC architecture enhanced with a high-performance GPU, while the other is based on a smartphone-style system-on-chip with dedicated AI acceleration. Some PC manufacturers have already begun shipping powerful NVIDIA GPUs since 2019, while Apple has defined the second option with its Apple Silicon MacBooks.
Omdia researchers developed various user personas to understand the different needs and performance requirements of users. The “Corporate Warrior” persona, which represents business users who heavily rely on collaboration tools and travel frequently, is the largest but least served by the current dominant players. Intel’s Core Ultra CPUs are targeting this persona to break back into the AI space, while Arm-based systems like Apple’s have an advantage with unified memory.
Unified memory is crucial for AI on PCs, as it impacts the type of AI models that can be run and the performance level. Arm’s investment in its unified memory architecture since the 1980s gives it a significant edge in this space. The Omdia report covers technology, user personas, major silicon vendors, and PC OEMs and is available in Omdia’s Advanced Computing Intelligence Service.
Omdia is a technology research and advisory group within Informa Tech, providing actionable insights for organizations to make smart growth decisions. For more information, contact Fasiha Khan at [email protected].