US Company Adjusts to Market Shifts, Aims to Regain Share in Crucial Market

Nvidia, a leading US-based artificial intelligence (AI) chipset manufacturer, is set to launch a new AI chip for China at a significantly lower price than its restricted H20 model. The move comes as the company seeks to adapt to the changing market landscape and regain its share in the crucial Chinese market.

The new chip, which is part of Nvidia’s latest generation AI processor with Blackwell architecture, is expected to be sold for between $6,500 and $8,000, a substantial reduction from the H20 price range of $10,000 to $12,000. According to sources, the lower price reflects the chip’s weaker specs and simpler manufacturing needs.

The new chip will be based on the Nvidia RTX Pro 6000D, a server-class graphics processor, and will use conventional GDDR7 memory, unlike the more sophisticated high-bandwidth memory used in the H20 model. Additionally, the chip will not employ advanced Chip-on-Wafer-on-Substrate (CoWoS) technology from Taiwan Semiconductor Manufacturing Co (TSMC), a decision that was likely made to comply with export restrictions.

Nvidia’s decision to launch a cheaper AI chip for China is a response to the company’s declining market share in the country. According to CEO Jensen Huang, Nvidia’s market share in China dropped dramatically from 95% before 2022, when US export restrictions came into effect, to 50% at present. Huawei has emerged as a major competitor, with its Ascend 910B chip gaining significant traction in the Chinese market.

The launch of the new AI chip is also seen as a strategic move by Nvidia to regain its footing in the Chinese market, which accounts for 13% of the company’s sales. However, the move may also be seen as a response to the ongoing export restrictions imposed by the US government, which have limited Nvidia’s ability to access the Chinese market.

“Nvidia’s market share in China dropped dramatically from 95% before 2022, when US export restrictions came into effect, to 50% at the moment,” Huang told reporters in Taipei this week. “If US export restrictions continue, more Chinese subscribers will switch to Huawei chips.”

The new export regulations, which introduce new limits on the metric GPU memory bandwidth, have also had a significant impact on Nvidia’s business. The company has been forced to record a supply loss of $5.5 billion, and Huang has warned that the company may also face a potential sale of $15 billion if the restrictions continue.

According to Jefferies’ investment bank, the new regulation limits memory bandwidth to 1.7-1.8 terabytes per second, compared to the H20’s ability to reach 4 terabytes per second. GF Securities projects that the new GPU will reach around 1.7 terabytes per second using GDDR7 memory technology, just according to export control limits.

The restrictions have also led to a significant shift in the Chinese market, with Huawei emerging as a major player. Huawei’s Ascend 910B chip has gained significant traction in the Chinese market, and the company has been able to capitalize on Nvidia’s decline.

The launch of the new AI chip is expected to be a crucial test for Nvidia’s ability to adapt to the changing market landscape and regain its share in the Chinese market. If successful, the move could help Nvidia regain its footing in the Chinese market and potentially limit the impact of the ongoing export restrictions.

However, the move may also face significant challenges, including the ongoing export restrictions and the increasing competition from Huawei. The company will need to carefully balance its pricing strategy with the need to maintain profitability and competitiveness in the market.

As the US-China trade tensions continue to escalate, the launch of the new AI chip is a significant development in the ongoing saga. The move highlights the challenges faced by US tech companies in accessing the Chinese market and the need for companies to adapt to changing market conditions.

The launch of the new AI chip is expected to take place in the coming months, with mass production scheduled to start as soon as June. The move is expected to have significant implications for the Chinese market and the global AI industry as a whole.