close
close

Chinese engineers secure access to NVIDIA AI chips through a new GPU rental service network

Chinese engineers secure access to NVIDIA AI chips through a new GPU rental service network

Chinese engineers are reportedly circumventing US restrictions by accessing NVIDIA’s AI chips through decentralized GPU rental services – a method that is gaining traction amid rising geopolitical tensions.


Chinese engineers circumvent US restrictions


China continues to circumvent the restrictions imposed by the Biden administration on the country’s access to advanced artificial intelligence.


Chinese engineers and companies are increasingly turning to “rental services,” which we will discuss later, to take advantage of Team Green’s processing capacity. This is an unexpected development.


NVIDIA AI chips fetch high prices


There were rumors that NVIDIA’s H100 would fetch exorbitant prices on China’s black markets, but clever engineers in the country have found a way around this.


According to the Wall Street Journal, Chinese engineers are working with certain international brokers to gain access to powerful computers and are using cryptocurrencies to hide their identities.


AI chips rented, not bought


NVIDIA’s high-end AI accelerators are still available in the Chinese market, but people prefer to rent them rather than buy them because the purchase price is much higher than the cost of a short-term rental.


A Bitcoin miner named Derek Aw has reportedly convinced Chinese consumers who want to rent computing capacity. This is because Aw convinced investors to set up massive AI clusters using NVIDIA’s H100 AI chips.


Taking advantage of the huge demand in the markets, the entrepreneur has reportedly set up 300 servers in a data center in Brisbane, Australia, and plans to further expand his rental business.


Large technology companies are joining the trend


Interestingly, this business model is being adopted by major technology companies.


As mentioned, CSPs like Microsoft, Google, and others are renting out servers based on NVIDIA’s A100 and H100 architecture to customers in China. The “GPU decentralized” industry has grown tremendously in the last two years as a result of this trend, which allows a limited number of countries to access high-end computing power, and it’s true that this does not currently violate any US laws.


WCCFTECH reports that Aw is also currently seeking funding to integrate NVIDIA’s Blackwell AI architecture into its offerings, giving customers access to the most advanced products on the market.


We don’t know the pricing plans, but we believe that GPU decentralization still has a long way to go due to the many wars in the world, especially between China and the United States.

Leave a Reply

Your email address will not be published. Required fields are marked *