ADVERTISEMENT
In the realm of AI (Artificial Intelligence), the United States of America has long dominated with OpenAI's ChatGPT, followed by Google, Meta and more. Meanwhile, Chinese AI platform DeepSeek V-3, launched a few weeks ago, created a buzz as its downloads jumped past ChatGPT on Apple’s App Store download rankings in the US, according to the Spectator Index.
The highlight of DeepSeek V-3’s development is that it was reportedly created at a fraction of the cost involved in developing OpenAI, Meta, etc, at just $5.6 million, which is very less as compared to the expenses involved in developing the others.
This is concerning for the US, as on his second day in office, President Donald Trump unveiled the Stargate Project, a groundbreaking $500 billion initiative bringing together tech giants OpenAI, Oracle and Softbank. Moreover, Trump emphasised the US’s commitment to staying ahead of China.
DeepSeek
Developed by the Hangzhou-headquartered startup, DeepSeek R1 is designed for reasoning and analytical tasks. It features a hybrid structure that combines reinforcement learning and chain-of-thought reasoning. It is available in two versions— DeepSeek-R1 and DeepSeek-R1-Zero.
While US-based AI companies tapped into the power of superior technology NVIDIA H100 GPUs, DeepSeek took a less powerful route of leveraging lesser capable H800 GPUs.
DeepSeek co-founder and CEO Lian Wenfeng reportedly purchased thousands of NVIDIA GPUs (Graphics Processing Unit) in 2021. In 2023, he introduced DeepSeek to focus on Artificial General Intelligence.
Earlier, Wenfeng shared that he was interested in working with PhD students from China’s premier universities.
Fortune India is now on WhatsApp! Get the latest updates from the world of business and economy delivered straight to your phone. Subscribe now.