How DeepSeek V3 Chinese AI Rival to ChatGPT Gemini Overcame America Chip Restrictions Build Cutting Edge System with Minimal Resources

Chinese startup DeepSeek has introduced its AI system, DeepSeek-V3, which will compete with OpenAI and Google’s popular and highly advanced AI chatbots (ChatGPT and Gemini respectively). What’s special here is that DeepSeek achieved this using far fewer computing resources, reportedly relying on only 2,000 Nvidia chips and about $6 million worth of computing power, compared to Meta’s $60 million. This innovation was inspired by US export restrictions on advanced chips, which prompted Chinese researchers to optimize training methods. DeepSeek-V3 can perform tasks such as question answering, logic problem-solving, and code generation. The company kept it open-source, which could also promote global cooperation in AI development.

Despite US efforts to limit China’s access to advanced technology, DeepSeek turned the challenge into an opportunity. The company has developed DeepSeek in this way with the help of many times less resources. developed It means that it can compete directly with the AI ​​products of OpenAI, Google, Microsoft and Meta. of The New York Times AccordingDeepSeek-V3’s engineers used far fewer specialized computer chips than their American competitors – which are essential for training the AI. These chips have been the main focus of the technological race between the US and China. The US imposed several restrictions on exports to maintain its AI lead, but the New York Times says DeepSeek’s success shows that Chinese researchers have become more innovative with the resources they have.

DeepSeek-V3 reportedly performed well at tasks such as answering questions, solving logic problems, and generating code. Even more impressive is the cost, with reports saying DeepSeek spent around $6 million on computing power, while Meta reportedly used $60 million for its latest AI product. While leading AI companies often use supercomputers with 16,000 or more chips, DeepSeek reportedly uses only 2,000 Nvidia chips.

Not only this, DeepSeek can become more efficient in the coming times, because the developers have adopted the open-source path with it. DeepSeek has open-sourced its latest AI system, sharing the code with others. This fosters global collaboration, as noted Yineng Zhang, a software engineer at Bestain.