Chinese Chip Designers, Including Tencent, Challenge Nvidia’s Dominance Amid US Restrictions


In the wake of heightened US export restrictions, Chinese chip designers, including major player Tencent Holdings, are aggressively positioning their AI chips as viable alternatives to those offered by Nvidia. Nvidia currently commands up to 90% of China’s $7 billion market for chips used in processing vast amounts of data for artificial intelligence (AI) software development.

The increased strategic technology controls imposed by the US have emboldened smaller players like state-backed Hygon Information Technology and startup Iluvatar CoreX to challenge Nvidia’s dominance. Huawei Technologies has emerged as a prominent contender, with its Ascend 910B drawing comparisons to Nvidia’s A100 in computing power, if not overall performance.

Despite these challenges, Tencent and other smaller AI companies are intensifying their efforts by accelerating chip product launches and increasing marketing initiatives. While the US rules primarily affect the most advanced chips, there is a belief that clients may still be swayed away from Nvidia.

Tencent, a leading social media and gaming firm in China, is actively promoting its AI inference chip, Zixiao, developed in collaboration with deep learning startup Enflame. Tencent emphasizes Zixiao’s performance, which it claims is comparable to some of Nvidia’s chips. While Tencent uses Zixiao chips internally, it rents computing power to external clients through its cloud services, offering Nvidia or AMD chips as alternatives.

In addition to direct sales, Tencent-backed Enflame and Iluvatar CoreX, among others, are marketing upcoming upgrades of their offerings as substitutes for Nvidia’s advanced A100 chip. Hygon is marketing its Shensuan No. 2 GPU, designed to be compatible with Nvidia’s CUDA computing platform.

The export curbs have created opportunities for Chinese tech firms to explore alternatives to Nvidia. While the potential financial impact on Nvidia is significant, especially regarding annual profits generated from Play Store commissions, concerns over production capacity may arise due to US restrictions on foundries working with Chinese firms.

Notably, the restrictions have stimulated a shift in strategy among tech giants in China, emphasizing the importance of having a variety of AI chips in stock rather than relying solely on Nvidia’s offerings. This shift has paved the way for Chinese cloud giants to build their AI ecosystems without depending on US chips, contributing to the development of China’s self-reliance in the AI sector.

Nvidia CEO Jensen Huang sees Vietnam as a potential second home for the company, planning to open a legal entity in the Southeast Asian country. Huang revealed plans for a design center and mentioned an existing investment of around $250 million in Vietnam.

Leave a Reply

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

Previous Post

U.S. Court Finds Google Maintained Monopoly Through Anticompetitive Practices

Next Post
google ai

Samsung Set to Integrate Google Pixel-Like AI Features in Galaxy Smartphones

Related Posts