State-owned China Telecom claims to have trained two LLMs—one with a 100-billion parameter and another with one trillion—purely using homegrown AI accelerators.

The Institute of AI at China Telecom revealed that their open-source TeleChat2-115B and another, unnamed model were trained using tens of thousands of domestically produced chips.

The state-owned telecoms operator did not reveal what chips it used for its 1 trillion-parameter model, but it has a partnership with Huawei.

As first reported by Reuters, ByteDance, the Chinese parent company of TikTok, is planning to train and develop an AI model using chips from fellow Chinese company Huawei. Three anonymous sources approached Reuters with this information;

TikTok's Chinese parent ByteDance plans to develop an AI model trained primarily with chips from compatriot Huawei Technologies, said three people familiar with the matter, as U.S. curbs turn the social media giant homeward in search of chips.

ByteDance (BDNCE) is reportedly planning to develop an artificial intelligence model, which will be trained mainly with chips from Huawei Technologies. The TikTok owner is now expected to use Huawei's Ascend 910B chip to train a large-language AI model,