A big Chinese telecom company, China Telecom, recently announced that it has created two new artificial intelligence (AI) models using chips made entirely in China. This is a significant step in China’s mission to become self-sufficient in technology, especially as the United States tightens its rules on selling advanced chips to China.
What Are These New AI Models?
The models, named TeleChat2-115B and another unnamed model, have been trained using tens of thousands of chips made in China. This achievement shows that China is making serious progress in AI, particularly in developing large language models (LLMs) that can understand and generate human-like text—similar to OpenAI’s ChatGPT.
According to the Institute of AI at China Telecom, this success indicates that China has reached a new level of self-reliance in AI technology. The unnamed model boasts an impressive 1 trillion parameters, which are like tiny switches that help the AI learn and make decisions. The TeleChat2-115B model has over 100 billion parameters. The more parameters an AI model has, the better it can perform tasks.
The Impact of US Restrictions
China’s drive to develop its own chips comes amid increasing restrictions from the US on advanced semiconductor technology. Companies in China are feeling the pressure and are working hard to find alternatives to American-made chips, like those from Nvidia. China Telecom is believed to be using chips from Huawei, a major Chinese tech company, to train its AI models.
Despite the challenges posed by US sanctions, Huawei is stepping up its game. The company is testing new AI chips and offering them to big tech firms in China as alternatives to Nvidia’s high-end chips. By using its own technology, China hopes to reduce its reliance on foreign suppliers and strengthen its position in the global AI race.
The Race for AI Supremacy
Chinese companies are in a race to catch up with global leaders in AI. With the US putting limits on chip exports, many companies are looking for ways to work around these restrictions. For example, Huawei’s Ascend chips are becoming popular among Chinese firms that need powerful hardware for their AI projects.
This push for homegrown technology is crucial as the AI industry grows. AI models need powerful chips to train effectively, and as Nvidia’s products become harder to access, Chinese firms are finding solutions closer to home. China Telecom is also considering chips made by another Chinese startup, Cambricon, to support its AI efforts.
Conclusion: A New Era for Chinese Technology
With these advancements, China is not only demonstrating its ability to develop sophisticated AI technologies but also signaling a shift in the balance of power in the tech world. As the country continues to innovate and invest in homegrown technology, it could soon emerge as a significant player in the global AI landscape, challenging established powers like the United States.
This development is just the beginning. China’s journey toward technological independence is unfolding, and the world is watching closely to see how it will impact the future of AI and the global technology race.
