crossorigin="anonymous"> US chip export rules suggest limits to thwart Chinese GPUs. – Subrang Safar: Your Journey Through Colors, Fashion, and Lifestyle

US chip export rules suggest limits to thwart Chinese GPUs.


The federal government on Jan. 13 proposed a global policy to limit the distribution of powerful, American-made GPUs, an effort to bolster American dominance over China in the AI ​​chip industry.

“Increasing U.S. national security and economic power requires that we not keep this critical technology offshore and that the world’s AI runs on American rails.” White House announcement reads

The White House went ahead with the proposal despite pushback from several tech giants, including NVIDIA and Oracle. If the rule is implemented, it will take effect 120 days after the comment period. The incoming administration will determine whether this rule will be enforced.

The proposed rule divides countries into allied, restricted or restricted categories.

The interim final rule on artificial intelligence proliferation proposes restrictions on U.S. chip distribution to protect U.S. market share against competing technologies in China. Issued under the Commerce Department’s Bureau of Industry and Security, the rule mitigates national security risks posed by AI, including Cyber ​​attacks.

The principle has six main parts:

  • 18 US allies will have unlimited access to chip sales.
  • The White House said that “chip orders with collective computation power of up to approximately 1,700 advanced GPUs” are exempt. Therefore, most universities and medical and research institutions can place chip orders without any hassle.
  • Trusted partners known as “Universal Verified End Users” have access to one of their global AI computational capabilities globally. The US is allowed to keep a small percentage.
  • Countries that are neither a U.S. “country of concern” nor on the list of close allies that meet certain security requirements may be listed as “nationally certified end users.” The designation allows them to purchase computational power equivalent to 320,000 advanced GPUs from the United States for use within neutral countries over the next two years. (“Countries of Concern” are China and its Special Administrative Regions, Hong Kong and Macau.)
  • Foreign governments, health care providers, and businesses outside the United States or its close allies not designated as national certified end users may purchase the equivalent of 50,000 advanced GPUs from the United States per country.
  • Governments that sign arrangements with the US for export control, clean energy, and technology protection efforts can double the number of chips they are allocated.

China will generally be held back by advanced US technology and AI foundation model weights. Russia is also barred from buying advanced chips from the US.

Which countries have unlimited access to USAI chips?

There are 18 allied countries with unlimited access to US AI chips under the policy:

  • Australia
  • Belgium.
  • Canada.
  • Denmark
  • Finland.
  • France.
  • Germany.
  • Ireland
  • Italy.
  • Japan.
  • The Netherlands.
  • New Zealand.
  • Norway
  • Republic of Korea (South Korea).
  • Spain.
  • Sweden
  • Taiwan.
  • United Kingdom.

As CNN Categories of countries that have caps on how much U.S. chip technology they can access, he says, can help prevent China from acquiring U.S.-made chips or gaining a foothold in the advanced technology industry in those countries. Israel and Mexico are among these countries. Limited access.

Depending on the size of their operations, companies outside the US may see delays in their supply chains due to the policy or limited opportunities to add AI features to their products.

“Although administration orders and export controls create a narrow, short-term advantage, they can cause broad, long-term damage to American technological leadership,” said Benjamin Lee, a professor of engineering and computer science at the University of Pennsylvania. told TechRepublic in an email. .

“In the short term, export controls will slow the deployment of the latest GPUs and largest AI data centers in some countries. But in the long term, exports Controls would cause other countries to develop their own hardware architectures or software models.

NVIDIA, Oracle strongly opposes this move.

Ned Finkel, NVIDIA’s vice president of government affairs, disagrees with the new rules on two key points. They argue that the distribution of powerful AI chips is a matter of national security and that banning the purchase of AI in some countries would benefit the US.

“Under the guise of an anti-China move, these rules will do nothing to enhance U.S. security,” Finkel wrote. “The new rules will govern technology around the world, including technology that is already widely available in mainstream gaming PCs and consumer hardware.”

See: EU approved. A big merger between simulation software company Ansys and chip design software provider Synopsys.

Ken Gluck, executive vice president at Oracle, acknowledged the importance of certain restrictions surrounding the use of AI in matters such as weapons of mass destruction. Super intelligent AI However, he also opposes the White House’s new policy, calling the rule “too complicated and wildly overbroad.” In a blog post on January 5.

“In one ambiguous action, BIS retroactively regulates global cloud GPU deployments. Shrinks the global market for US cloud and chip suppliers; establishes volume restrictions; tells 20 countries that they can only be trusted if they agree to unilaterally imposed new conditions — including certification and semi-annual reporting requirements — and potentially push the rest of the world toward Chinese technology,” Gluck said. wrote

however, In a statement to The Associated PressMicrosoft president Brad Smith was less critical: “We are confident that we can fully comply with the high security standards of this rule and meet the technology needs of countries and customers around the world who rely on us.” are.”



Source link

Leave a Reply

Translate »