SINGAPORE/BEIJING (Reuters) – Nvidia’s Chinese customers have been informed that the launch of a new artificial intelligence (AI) chip, the H20, designed to comply with U.S. export rules, is being delayed until first quarter of next year, according to two sources familiar with the matter.

The H20 is the most powerful of three China-focused chips that Nvidia has developed to comply with new U.S. export restrictions, the sources said. The delayed launch could complicate the Santa Clara, California-based group’s efforts to preserve its market share in China against competitors such as Huawei.

Nvidia was expected to launch its new products as early as November 16, as industry newsletter SemiAnalysis reported this month.

The H20 launch was ultimately pushed back to the first quarter of next year, the sources said, with one saying they had been told the launch could happen in February or March.

Both sources declined to be named as the information is confidential. Nvidia declined to comment.

The sources said they were informed that the H20 launch was being delayed due to problems server makers faced integrating the chip.

In addition to the H20, Nvidia planned two other chips to comply with new US export rules: the L20 and the L2. The sources said the launch of the L20 was not delayed. They were not able to provide information on the progress of the L2.

Nvidia is banking on the chips to preserve its market share in China after Washington’s toughened export rules barred it from shipping products such as the A800 and H800 AI chips.

(Reporting Fanny Potkin in Singapore, Yelin Mo in Beijing and Brenda Goh in Shanghai; Augustin Turpin, edited by Blandine Hénault)

Copyright © 2023 Thomson Reuters