MediaTek Unveils 2nm Dimensity 9600 Pro AI Chip
MediaTek is pushing its premium smartphone platform into the 2-nanometer era with the Dimensity 9600 Pro, its first flagship 5G Agentic AI chip built on Taiwan Semiconductor Manufacturing Co.’s 2nm process. The processor targets top-tier handsets as device makers race to run more generative AI workloads locally. More efficient computing and memory use are becoming critical differentiators as increasingly large models move from cloud servers onto smartphones.
The newly unveiled Dimensity 9600 Pro uses a redesigned scheduling engine and NPU model-compression technology to reduce the memory required for AI workloads. MediaTek said the chip can support 30-billion-parameter mixture-of-experts, or 30B MoE, models for efficient on-device processing, while delivering a 51% increase in AI performance. The first smartphones powered by the processor are scheduled to be announced in September, with devices expected to reach Taiwan as early as October.
All Coverage
2 original reportsThe Backstory
The history behind this eventMediaTek Posts Record August Revenue as Nvidia Deepens AI Ties
MediaTek, one of the world’s largest fabless chip designers, is expanding beyond its smartphone base into AI PCs and custom data-center silicon. Its next flagship mobile processor, the Dimensity 9600, is expected to use Taiwan Semiconductor Manufacturing Co.’s 2-nanometer process and support on-device Agentic AI, underscoring the industry’s push to run more generative-AI workloads locally rather than relying solely on the cloud.
MediaTek’s consolidated revenue exceeded NT$64.1 billion in August, setting a monthly record. Nvidia also invested $3.5 billion through a strategic subscription to convertible bonds, strengthening cooperation across AI PCs and the cloud ASIC computing ecosystem. The revenue milestone, upcoming Dimensity 9600 launch and deeper Nvidia relationship prompted foreign brokerages to raise their ratings and price targets, with the most bullish target reaching NT$6,800.
Meta, Broadcom Extend AI Chip Partnership Through 2029, Plan 2-Nanometer MTIA Deployment
Meta has long worked with Broadcom to develop its in-house MTIA AI accelerator chips. Custom designs for recommendation, advertising and generative AI workloads improve data-center computing efficiency. The initiative is central to securing the computing power Meta needs to build “personal superintelligence” and helps reduce its reliance on external GPU suppliers such as NVIDIA, along with the associated supply risks.
Meta and Broadcom have extended their MTIA partnership agreement through 2029 and plan to expand chip deployment. The collaboration will focus on the first MTIA chip built using a 2-nanometer process, which is expected to become a key foundation of Meta’s in-house AI computing infrastructure. The contract value and precise mass-production date have not been disclosed, but the agreement explicitly runs through 2029.
Subscribe to Mark Radar Weekly
Every Friday, the week's strongest signals in your inbox. Unsubscribe anytime.
If you search news on Google, you can set Mark Radar as a preferred source—our coverage will show up more often in your results. Set as preferred source on Google →