Drooid Logo
Back to story perspectives

Full Breakdown

Arm Unveils C2 CPU Cores and Mali G2-Ultra NX GPU for Next-Gen Android Chips

9/9/2026, 8:51:56 AM

Announcement Overview

Arm announced a new family of C2 CPU cores—C2-Ultra and C2-Pro—alongside an upgraded Mali G2-Ultra NX graphics processor. The company said the designs will appear in upcoming Android smartphones, with the first implementation slated for Xiaomi’s 18 Fold device that uses the XRing O3 chip.

Technical Details and Performance Claims

Arm describes the C2-Ultra and C2-Pro cores as delivering a 15 percent increase in single-thread performance and a 12 percent boost in multi-thread throughput at the cluster level. The company also cites a 12 percent reduction in application launch time and a 15 percent faster web-browsing experience. For artificial-intelligence workloads, Arm claims up to a 1.7-times performance improvement when models run on the CPU’s built-in AI engine, eliminating the need for a separate NPU or TPU.

The Mali G2-Ultra NX GPU is positioned as “desktop-class” for mobile, with an overall 14 percent performance uplift. Arm highlights the integration of “AI-native graphics,” noting that neural acceleration is embedded directly in the graphics pipeline. The GPU supports Unreal Engine 5 “MegaLights” and can generate intermediate frames to smooth gameplay or upscale lower-resolution content.

Integration into Devices

Arm’s designs are already used in many Android chipsets, including MediaTek flagship SoCs and Google’s Tensor processors. The upcoming Xiaomi 18 Fold, powered by the XRing O3 chip, will be the first phone to combine the new C2 CPU cores with the Mali G2-Ultra NX GPU.

Potential Impact on Mobile Gaming and AI

By merging AI acceleration with both CPU and GPU functions, Arm aims to enable richer visual experiences and more responsive applications while staying within mobile power limits. The ability to upscale games and render advanced lighting effects could narrow the gap between mobile and desktop gaming performance. Additionally, on-CPU AI processing may simplify device designs and reduce latency for AI-driven apps.