The Chinese corporation Alibaba has managed to develop its own artificial intelligence chip, which is capable of filling the deficit in the Chinese market due to sanctions restrictions on Nvidia.
The Chinese corporation Alibaba has managed to develop its own artificial intelligence chip, which is capable of filling the deficit in the Chinese market due to sanctions restrictions on Nvidia.
This is reported by Wccftech with reference to WSJ. It is claimed that the chip for Alibaba’s artificial intelligence is designed to handle workloads and may be created using SMIC’s 7nm process technology.
China has previously been moving away from American AI technology and focusing on domestic solutions. Companies such as Huawei, Cambricon, MetaX, and many others are increasing their investments in developing their own chips. Alibaba is said to have become one of the first to make its own AI chip, capable of filling the niche of Nvidia products.
Alibaba’s chip is said to have been developed by a Chinese company, likely SMIC, and is currently being tested on workloads related to summarization. SMIC currently has 7nm production lines, so it is likely that Alibaba used this process to produce its AI chip, putting it on par with solutions from Huawei and Cambricon.
China still faces the problem of a shortage of production capacity as its factories cannot support large-scale production, mainly due to old equipment such as DUV lithography and inefficient technology.
Chinese companies still need NVIDIA AI chips to train models, and so far no other option can replace American technology to build powerful LLMs. Furthermore, Nvidia CEO Jensen Huang insists on creating a Blackwell-based product for China, indicating a pressing need for the company to develop a solution to stay relevant in this large market.
In April of this year, it was reported that the Donald Trump administration had banned Nvidia from exporting its new H20 chip to China, even though it was specifically designed to comply with previous US restrictions. The company’s losses from this decision were estimated at $18-20 billion.
