×
New Research Explores Global Divide in AI Chip Ownership
Written by
Published on
Join our daily newsletter for breaking news, product launches and deals, research breakdowns, and other industry-leading AI coverage
Join Now

AI chip distribution reveals global digital divide: New research uncovers a stark imbalance in the global distribution of advanced AI chips, highlighting the concentration of computational power in a handful of countries and its potential impact on AI development and governance.

Key findings of the study: A peer-reviewed paper, shared exclusively with TIME, maps the locations of publicly accessible GPU clusters worldwide, revealing significant disparities in AI infrastructure distribution.

  • The United States and China lead in the number of GPU-enabled regions, with China having more overall but the U.S. dominating in the most advanced GPU availability.
  • Only 30 countries worldwide have significant GPU presence, leaving much of the world in “Compute Deserts” with no access to hire GPUs.
  • The U.S. has eight regions where the most advanced H100 GPUs are available for hire, while China has none publicly accessible.

Implications for AI governance and development: The concentration of AI infrastructure in specific countries raises concerns about global influence on AI development and regulation.

  • Countries hosting AI infrastructure have greater power to enforce compliance with their regulations and shape AI development norms.
  • Nations without local AI infrastructure may have limited legislative choices and less influence over AI’s future direction.
  • This disparity could lead to a situation where countries in the “Compute South” become reliant on AI systems developed in the “Compute North” without having input into their design or operation.

Limitations of the study: The research provides valuable insights but has some notable constraints.

  • The study only accounts for publicly accessible GPU clusters, excluding those owned by governments or private tech companies for internal use.
  • It doesn’t consider non-GPU chips increasingly used in AI applications.
  • The research counts compute “regions” rather than individual chips, due to the confidential nature of specific GPU distribution information.

Geopolitical implications: The findings underscore the growing importance of advanced chips in global competition and AI development.

  • The U.S. and China are engaged in a race to accumulate high-end chips, with the U.S. imposing sanctions to limit China’s access to cutting-edge varieties.
  • Despite sanctions, reports suggest a black market for restricted chips has emerged in China, with millions of dollars worth of chips allegedly smuggled into the country.

Global divisions in AI compute: The researchers propose a new framework for understanding global AI infrastructure distribution.

  • “Compute North” refers to regions with the most advanced chips, primarily suited for both training and running AI systems.
  • “Compute South” includes areas with older chips, capable of running but not training advanced AI systems.
  • “Compute Deserts” are regions with no publicly accessible GPU clusters.

Potential consequences of the divide: The concentration of AI infrastructure in wealthy economies could have far-reaching effects on global technological development and economic inequality.

  • This distribution mirrors existing patterns of global inequalities between the Global North and South.
  • It may reinforce the economic, political, and technological power of “Compute North” countries.
  • “Compute South” countries may face challenges in shaping AI research and development, potentially limiting their agency in the global AI landscape.

Analyzing deeper: The need for equitable AI development: The research highlights the critical importance of addressing the global AI infrastructure divide to ensure more inclusive and diverse AI development.

  • Efforts to democratize access to AI compute resources could help prevent the entrenchment of existing global inequalities in the AI era.
  • International cooperation and investment in AI infrastructure in underserved regions may be necessary to create a more balanced global AI ecosystem.
  • As AI continues to shape various aspects of society and the economy, ensuring broader participation in its development becomes crucial for addressing global challenges and promoting equitable technological progress.
Exclusive: Research Finds Global Divide in AI Chip Ownership

Recent News

Baidu reports steepest revenue drop in 2 years amid slowdown

China's tech giant Baidu saw revenue drop 3% despite major AI investments, signaling broader challenges for the nation's technology sector amid economic headwinds.

How to manage risk in the age of AI

A conversation with Palo Alto Networks CEO about his approach to innovation as new technologies and risks emerge.

How to balance bold, responsible and successful AI deployment

Major companies are establishing AI governance structures and training programs while racing to deploy generative AI for competitive advantage.