August 24, 2026 – Xiaomi has officially unveiled three new proprietary chips today: the Xuanjie O3, Xuanjie O100, and Xuanjie D100. Positioned respectively as an AI flagship SoC, a high-bandwidth AI acceleration chip, and a high-compute AI chip for autonomous driving, the trio has completed all validation phases. The Xuanjie O3 will debut in the upcoming Xiaomi 18 Fold flagship, while the O100 and D100 are slated for commercial release next year.
The simultaneous launch of these three chips prompted China Central Television (CCTV) to praise the achievement on Weibo, stating, “Three chips launched simultaneously mark another breakthrough in China’s chip industry, as Xiaomi’s Xuanjie series continues to explore a ‘full-ecosystem AI computing foundation.'” Xiaomi CEO Lei Jun emphasized that Xuanjie is no longer just a chip strategy but the physical foundation of Xiaomi’s AI strategy, spanning from pockets to car cabins, living rooms, and factories.
The standout of the launch is the Xuanjie O3, a next-generation flagship SoC built on a 3nm process. Packing 24 billion transistors into a compact 133mm² die, it represents a 26% increase over its predecessor, the Xuanjie O1. Performance-wise, the O3 achieved an AnTuTu benchmark score of 5.22 million, significantly outpacing current flagship SoCs. Its CPU features a 10-core all-big-core architecture with a peak clock speed of 4.35GHz, delivering a 31% single-core and 60% multi-core performance boost in Geekbench 6.5 tests. Lei Jun also expressed immense confidence in the chip’s graphics capabilities; the O3 debuts a 16-core G2-Ultra NX GPU with third-generation ray tracing, offering an 85% performance uplift and a 64% reduction in power consumption.
First introduced in May 2025, the original Xuanjie O1 was mainland China’s first 3nm flagship SoC and powered devices like the Xiaomi 15S Pro. The expansion from a single mobile SoC to a comprehensive three-chip matrix marks a pivotal moment for Xiaomi. The company’s proprietary chip efforts have now evolved to cover end-side AI computing demands across smartphones, AI acceleration, and smart vehicles, effectively laying the groundwork for its “Human-Car-Home” full-ecosystem strategy.
