Deepsek is a Chinese Startup that has brought a storm in the Global Tech World by releasing the Artificial Intelligence (AI) model, which can compete with the artificial intelligence models created by American giants like Google and Openai. The specialty of Deepseek is that it has developed AI models using less advanced and inexpensive computer chips, usually different from high-performance chips used by companies such as NVidia and Alphabet.
Deepsek, established by Chinese Developer Liang Wenfeng in 2023, spent less than $ 6 million to train its AI model, which is a small part of a multi -billion dollar budget by American companies such as Openai, Alphabet and Meta. Deepsek has given the perception challenge that there is a need for heavy investment to create advanced AI.
Deepsek, a game changer
The 2,000 Nvidia H800 chips were used to train the Deepseek-R1 model, which are considered less advanced than today’s most state-of-the-art chips. Despite this, the engineers of Deepsek used several special models to make it more efficient that were able to process data effectively. This proved that high-performance chips are not required for this and AI models can also be developed with affordable technology.
Deepsek challenging American Tech World
The emergence of Deepseek-R1 challenges the notion that the US is the leader in the AI, as American companies invest heavy in their research, talent and data centers due to major financial capacity. Deepsek has proved that competitive AI models can also be developed using limited budget and old chips, which has long been accepted in Tech World.
After the release of the Deepseek model, NVIDIA shares fell by 17%, causing a loss of about $ 600 billion. This shows that companies such as NVIDIA had a monopoly on production of chips in the field of AI, but are now facing competition.
What will happen to the future of Deepsek
Deepsek’s success becomes even more important in terms of tech war between American-China, where the US has imposed export restrictions on high-demonstration chips to prevent China’s AI progress. However, Deepseek showed that the most advanced chips are not required to create state -of -the -art AI models, and China still has enough chips available, which can be used in innovation.
In 2023, Liang reported that his company had deposited the stock of NVIDIA A100 GPU before the US government banned the export of these chips. This indicates that Chinese companies are still capable of receiving high-performance chips, and progress in AI may continue despite American sanctions.
China’s growing strength in AI
Deepsek is part of China’s rapidly growing AI industry, which is now giving a tough challenge to America. This proves that China is now equal to America in the field of AI, and in the coming times this tech war can be even faster. Although it is too early to say that China is winning the AI race, the success of Deepseek has shown that AII does not require billions of dollars to excel in AI. Small players can also be able to be able to have the right approach and resources.
