MediaTek SoCs to use Meta’s Llama 2 to handle gen AI processing at Edge

MediaTek developing technologies to enable generative AI capability in edge devices such as smart phones, IOT, automotives and smart homes. MediaTek is working on making Meta’s open-source Large Language Model (LLM) Llama 2 to run on its latest APUs and NeuroPilot AI Platform. The idea is to enable edge devices to perform Generative AI processing at edge without depending on the cloud computers, so that they can perform on-device AI processing at any time during network unavailability. The Other advantages include seamless performance, greater privacy, better security and reliability, lower latency and lower operation cost. Although many of the present SOC chips designed for the edge computers have a lot of AI capabilities, they are still not fully leveraged for effective AI applications running using on-device computer power due to lack of support to interface with AI software models available in the market which are basically designed for cloud environment. MediaTek said its 5G smartphone SoC shipped today is equipped with APUs designed to perform a wide variety of Generative AI features, such as AI Noise Reduction, AI Super Resolution, AI MEMC and more. MediaTek said its next-generation flagship chipset, to be introduced later this year to feature a software stack optimized to run Llama 2, as well as an upgraded APU with Transformer backbone acceleration, reduced ...