Xiaomi Xuanjie O3, Apple M6/M5 Ultra chips have been unveiled one after another! Arm officially sent congratulations.

This week, Xiaomi's Xuanjie O3 and Apple's M6/M5 Ultra chips were unveiled. As chip products built on the Arm architecture, Arm officially posted on X platform, congratulating both Xiaomi and Apple.

These two products represent flagship self-developed achievements of the Arm architecture in the mobile and desktop segments respectively, also reflecting Arm ecosystem's technical progress across different hardware tracks.

Arm stated that Xiaomi's Xuanjie O3, equipped with an Arm CPU accelerated by SME2 and paired with a new-generation Mali GPU, helps this latest flagship SoC enhance AI computing, general-purpose computing, and graphics rendering capabilities on next-generation devices.

The Xuanjie O3 is Xiaomi's highest-positioned AI flagship SoC, featuring a ten-core all-big-core CPU paired with a 16-core G2-Ultra GPU, with AnTuTu scores exceeding 5.22 million. It will be used in high-end phones and tablets such as the Xiaomi 18 Fold. The chip is deeply customized based on Arm IP, with targeted optimizations for high-load scenarios like on-device AI large models and large-scale gaming.

Among its features, the G2-Ultra also makes its debut here. Compared to the previous-generation GPU, it delivers a massive upgrade with impressive performance and efficiency—performance is up 85%, making it the strongest mobile graphics processing capability on the market, completely outperforming the A19 Pro in benchmark tests, while power consumption is reduced by 64% compared to the previous generation.

Regarding Apple's M6 and M5 Ultra, Arm commented that these two Arm-based chips mark a significant step forward in AI computing and performance.

The M6 is Apple's first 2nm process chip, equipped with a 12-core CPU, 12-core GPU, and dual 16-core Neural Engines, with unified memory bandwidth reaching 170GB/s, for the new Mac mini.

The M5 Ultra targets desktop workstations, primarily featured in Mac Studio devices, focusing on ultra-high-density AI compute output.