According to CEO Jensen Huang, Nvidia will keep working on creating new products that abide with US government rules on the export of premium chips to China. Huang made this announcement on Wednesday.
More than 90% of China’s $7 billion AI chip industry is controlled by the California-based artificial intelligence chip designer, but analysts predict that as US restrictions on chip exports tighten, Chinese competitors will have greater opportunity to gain ground.
At a press conference in Singapore, Huang stated, “Nvidia has been working very closely with the U.S. government to create products that comply with its regulations.”
“At this point, our strategy is to keep collaborating with the government to develop a new line of products that adhere to the new, limited regulations.”
In light of the new US regulations, the company issued a warning on its November results call that it anticipates a sharp decline in fourth-quarter sales in China.
According to sources familiar with the situation, Reuters revealed last month that Nvidia had informed Chinese clients it would postpone the release of a new AI chip it had created to abide by US export regulations until the first quarter of 2019.
Huang added on Wednesday that the business was in discussions with Singapore over possible large investments.
Yotta Data Services, a data centre owned by the Hiranandani group, announced on Tuesday that it intends to install more than 20,400 NVIDIA GPU-based supercomputers by June 2024.
GPUs (Graphics Processing Units) are the main chipsets used to handle data related to artificial intelligence technologies. GPUs are normally used to process multimedia data. On the other hand, these chipsets are in short supply worldwide.
NVIDIA is the industry leader at the moment, with a roughly 88% global market share.
Yotta Data Services said in a statement that it has teamed up with NVIDIA to supply state-of-the-art GPU computing platforms and infrastructure for its Shakti Cloud platform, which will be the fastest supercomputer in the country with 16 Exaflops of AI computing power.
The alliance would expedite the development of AI solutions in India by granting access to a broad spectrum of organisations, corporations, AI researchers, and startups across the country.
Furthermore, Yotta is putting into practice a reference architecture that is NVIDIA-powered and outfitted with NVIDIA InfiniBand networking. This will allow GPU clusters to deliver exceptional performance at scale for applications involving AI training and inferencing as well as large-scale HPC.
