Get the App
SLTechnology News&Howtos  ›  IT Information  › 

MediaTek confirms that with the help of the Tianji 9300 processor, the generative AI will run locally on the phone.

Shulou Source: shulou.com Published: 2023-11-24 18:57:10 11月16日 Update

CTOnews.com, August 23, Qualcomm announced last month that with Meta's Llama 2 language model (LLM), in-device generated AI will be used in 2024 flagship phones. Now, MediaTek has confirmed that it will also provide device-side generation artificial intelligence.

MediaTek revealed that it is working with Meta to better support the Llama 2 LLM on phones powered by the company's next-generation flagship processor (CTOnews.com 9300).

MediaTek said: "MediaTek's next flagship chipset, to be launched later this year, will further enhance LLM and AIGC performance with software stacks optimized for running Llama 2, as well as upgraded APU with Transformer backbone acceleration, reduced footprint access and DRAM bandwidth use."

MediaTek said it expected that generative artificial intelligence applications that support Llama 2 would be available for phones with its next-generation chipsets. The company added that the phones were scheduled to go on sale by the end of 2023.

Both MediaTek and Qualcomm have announced that better support for generative AI on devices is a good thing because running spanning AI tasks locally has a privacy advantage because users' data is not sent to the cloud for reasoning.

As for the Tianji 9300 chipset, it is reported that it will adopt a radical design. CTOnews.com noted that the company has confirmed that the chipset will use Arm's Cortex-X4 and A720 CPU cores, as well as Immortalis-G720 GPU, and it has been revealed that the chipset will use four Cortex-X4 cores and four A720 medium cores.

Tags: MediaTek generation chip chipset mobile phone artificial intelligence intelligence next generation company flagship device support Tianji operation technology Qualcomm application processor processing Apple Docker Huawei Linux macOS MariaDB Microsoft MySQL NVidia OPPO Reno Docker Redmi MariaDB NVidia Huawei