This marked the first time a Chinese developer publicly validated the feasibility of using only Huawei chips to train AI models with MoE architecture, which has become widely adopted because of its ability to deliver high performance with fewer computational resources.
The TeleChat3 models – ranging from 105 billion to trillions of parameters – were trained on Huawei’s Ascend 910B chips and its open-source deep learning AI framework MindSpore, according to a technical paper published last month by China Telecom’s Institute of Artificial Intelligence (TeleAI).
The Huawei stack was able to meet the “severe demands” of training large-scale MoE models across many different sizes, TeleAI researchers said.
“These contributions collectively address critical bottlenecks in frontier-scale model training, establishing a mature full-stack solution tailored to domestic computational ecosystems,” they said.
The initiative by China Telecom, one of the world’s largest fixed-line and mobile network operators, underscored growing efforts by a number of US-blacklisted Chinese firms – including Huawei and iFlytek – to train AI models using domestically designed semiconductors.
semiconductors, artificial intelligence, OpenAI, Mixture-of-Experts architecture, Zhipu AI, fintech, H200, iFlytek, AI models, Huawei, China, Beijing, China Telecom, US, TeleChat3 models, Alibaba Group Holding, TeleAI, MoE architecture, Ant Group, Nvidia, Tsinghua University, Ascend 910B chips, DeepSeek's V3 model, Advanced Micro Devices, MindSpore#China #Telecom #develops #countrys #MoE #models #trained #Huaweis #chips1768918715
Advanced Micro Devices
AI models
Alibaba Group Holding
Ant Group
artificial intelligence
Ascend 910B chips
Beijing
China
China Telecom
chips
countrys
DeepSeek's V3 model
develops
fintech
h200
Huawei
Huaweis
iFlytek
MindSpore
Mixture-of-Experts architecture
models
MoE
MoE architecture
nvidia
openai
semiconductors
TeleAI
TeleChat3 models
telecom
trained
Tsinghua University
US
Zhipu AI
