Separator

MediaTek Unveils AI Smartphone Chipset with Latest TSMC 3-nm Tech

Separator

Chip-maker MediaTek unveiled Dimensity 9400, the new flagship smartphone chipset optimized for edge-AI applications, immersive gaming, incredible photography, and more. The first smartphones powered by the Dimensity 9400 chip will be available in the market, starting in Q4, said the company.

The Dimensity 9400, the fourth and latest in MediaTek’s flagship mobile SoC lineup, offers a boost in performance with its second-generation ‘All-Big Core’ design built on Arm’s v9.2 CPU architecture, combined with the most advanced GPU and NPU for extreme performance in a super power-efficient design.

Joe Chen, President at MediaTek, said the new chip will continue expanding “our mission to be the enablers of AI, supporting powerful applications that anticipate users’ needs and adapt to their preferences, while also fueling generative AI technology with on-device LoRA training and video generation”.

Built on TSMC’s second-generation 3nm process, the Dimensity 9400 is up to 40 percent more power-efficient than its predecessor, according to the company, also allowing users to enjoy longer battery life The Dimensity 9400 offers 35 percent faster single-core performance and 28 percent faster multi-core performance compared to MediaTek’s previous generation flagship chipset, the Dimensity 9300.

“As the fourth-generation flagship chipset, the Dimensity 9400 continues to build on our momentum of steady growth in market share, and MediaTek’s legacy of delivering flagship performance in the most efficient design for the best user experiences”, Chen added.

To allow users to take advantage of the latest generative AI applications, the Dimensity 9400 offers up to 80 percent faster large language model (LLM) prompt performance while also being up to 35 percent more power efficient than the Dimensity 9300. The company said it is working with developers to offer a unified interface between AI agents, third-party APKs, and models that efficiently run both edge AI and cloud services.

Current Issue