Nvidia enhances hardware optimization for Chinese open AI models, warning about US regulatory risks
- Nvidia is optimizing chips for Chinese open AI models.
- The company warns of potential U.S. restrictions.
- Tech giants are against premature sanctions.
- Chinese chips are becoming a competitor in developing countries.
The semiconductor giant Nvidia announced the launch of a program aimed at enhancing the efficiency of its hardware when working with open AI models created in China. As part of the initiative, the company stated its support for projects such as DeepSeek, Qwen from Alibaba, as well as its own and Google models.
In a document submitted to the Securities and Exchange Commission, Nvidia indicated that it is considering potential risks associated with possible interference from the U.S. government in the use of technologies developed in China. The firm emphasized that any restrictions could limit its ability to offer products and services built on these models.
According to industry analysts, by 2026, Chinese open AI models will reach a level comparable to leading Western solutions and are already being used by developers worldwide, including in the U.S. This growth raises concerns among lawmakers, who see it as a threat to U.S. technological leadership.
Measures are being discussed in Washington, ranging from the introduction of new tariffs on American semiconductors to possible restrictions on support for Chinese AI applications. A White House representative noted that there are currently no official decisions, and reports about tariffs are speculative.
A Nvidia representative emphasized that global support for models allows developers to utilize American technological infrastructure. According to him, if developers choose a stack optimized for a specific model, ensuring such optimization becomes critically important for maintaining the competitiveness of the U.S.
Tech companies oppose premature restrictions
In addition to Nvidia, more than twenty major players, including Microsoft, Meta, and Palantir, have joined the initiative. They issued an open statement urging lawmakers to refrain from 'premature' measures that could stifle the development of open-source in the AI field.
Chinese chip manufacturers, such as Huawei with its Ascend processor line and Alibaba, have also announced their own optimizations for open models, including DeepSeek and Qwen. This indicates a growing interest in self-deployment and modification of AI technologies without reliance on closed solutions.
Experts warn that more affordable Chinese solutions could become the de facto standard in developing countries, where the cost of licenses for Western systems remains high. In such a scenario, competition for technological superiority may enter a new phase, where open source and local optimization will play a crucial role.
Source: 3DNews
