According to four individuals who are acquainted with conversations of this nature, Chinese chip makers such as Tencent Holdings (0700. HK) are actively marketing their artificial intelligence processors as contenders to Nvidia’s (NVDA.O) chips in the hope that export limitations imposed by the United States will encourage customers to switch.

Nvidia, a California-based company, controls about 90% of China’s seven billion dollar market for processors used to handle enormous amounts of data to create artificial intelligence (AI) software.

On the other hand, the United States’ strategic technology regulations, which became more stringent in October, have given even less well-known companies, such as the state-backed Hygon Information Technology (688041. SS) and the upstart Iluvatar CoreX, the confidence to challenge the dominant player in the United States.

With its Ascend 910B, Huawei Technologies (HWT.UL) is being compared against Nvidia’s A100 in terms of computational power but not overall performance. With this comparison, Huawei Technologies is primarily seen as making the most significant progress.

According to the sources, Tencent and other smaller AI firms are speeding up the debut of their chip products and organizing more marketing trips. They are wagering that even if the laws in the United States only affect the most sophisticated chips, they might still cause customers to stop purchasing Nvidia products.

According to two individuals, Tencent, China’s largest social media and gaming company that provides cloud services to external clients, has been promoting services that use the artificial intelligence inference chip Zixiao, which it created in collaboration with the deep learning startup Enflame. These services boast performance that is equivalent to that of some Nvidia processors.

According to one of the individuals, Tencent is marketing its Zixiao v1 model as a more affordable alternative to Nvidia’s A10, which is utilized for artificial intelligence applications that involve picture and speech recognition. Additionally, it is promoting a forthcoming iteration of the v2Pro optimized for artificial intelligence training capable of replacing Nvidia’s L40S, which is now prohibited.

Tencent does not sell Zixiao chips directly to customers outside the company; instead, it utilizes them internally. It allows clients to rent computer capacity through its cloud services, which also provide access to Nvidia or AMD (AMD.O.) CPUs as alternatives.

Tencent has said it does not intend to develop Zixiao further than the currently available version. Tencent created the Zixiao platform to complement our cloud-based products and services while abiding by all applicable laws and regulations. It is only accessible to customers who have purchased Tencent Cloud’s enterprise services, according to a spokeswoman for the cloud computing giant.

Other people are pitching direct sales. According to two of the people, Enflame, which receives funding from Tencent and produces the Tiangai graphics processing unit (GPU), and Iluvatar CoreX, which makes the Yunsui artificial intelligence training accelerator chip, have both been marketing upcoming improvements of their products as potential substitutes for Nvidia’s potent A100 processor.

According to a third party, Hygon is selling its recently announced graphics processing unit (GPU), Shensuan No. 2, as being developed from the beginning to be compatible with Nvidia’s chip computing platform, CUDA. This meant that customers of Nvidia chips could move to a different chip with just minor design modifications.

To compete with the upcoming H20 chip, which Nvidia developed by the most recent export restrictions, Intellivision, a startup company, announced the Deepedge10 processor.

Intellivision allegedly rushed its launch to take advantage of the scenario Nvidia is currently facing. The company is licensing the chip to customers even though it has not yet been manufactured in large quantities.

Requests for comment were made to Enflame, Iluvatar CoreX, Hygon Information, and Intellifusion; however, none of these companies responded promptly. Nvidia declined to comment on it.

CAPABILITIES FOR PRODUCTION

Companies in the technology sector of the world’s second-largest economy have begun to look for alternatives to Nvidia. Tencent has stated that it was forced to hunt for domestic suppliers of artificial intelligence training chips in the United States. In addition, the most popular search engine on the internet, Baidu (9888. HK), reportedly made a significant order for Huawei chips a month ago.

We have many new businesses competing with us in China, all startups. There are around fifty startups that are centered on artificial intelligence. One of the most formidable competitors is Huawei. Other companies in the United States, such as Intel and AMD, are fierce competitors. Jensen Huang, the CEO of Nvidia, warned reporters in Singapore the previous week that the company faces significant competition.

Lucy Chen, Vice President of Isaiah Research, stated that even if Chinese chip designers were to receive orders, they could still have a difficult time finding production capacity due to the restrictions that the United States government imposes on foundries like TSMC (2330. TW) that prevent them from partnering with Chinese companies.

Huawei will likely be given priority for utilizing the bulk of China’s sophisticated manufacturing processes and advanced packaging capabilities. These up-and-coming businesses must devise a plan that will allow them to overcome the limitations and restrictions imposed on their production by the United States of America.

According to Nori Chiou, investment director at White Oak Capital, the restrictions have created a market opportunity. This is because tech companies are pursuing a strategy of having multiple types of artificial intelligence chips in stock rather than only Nvidia’s. Furthermore, the sustainability of their AI strategy is becoming a priority rather than performance.

“The United States’ original goal was to slow down China’s AI capabilities but, in fact, related action has boosted China’s self-development capability,” stated the president.

“Many Chinese cloud giants are working on building their AI ecosystems without U.S. chips due to these restrictions.”