Winning the silicon race: Three strategies to secure AI advantage

As companies step up efforts to capture AI-powered business opportunities, big ambitions are riding on cutting-edge, nanoscale semiconductors. But as global production and resource bottlenecks persist, there aren’t enough advanced AI chips to go around. How is this chip shortage impacting buyers and suppliers? A new report from the IBM Institute for Business Value, in partnership with SEMI, indicates that an already constrained AI semiconductor market is about to get even tighter. 

By 2028, 62% of chip buyers say high-performance AI computing infrastructure will be central to competitive advantage.

How fast is demand for AI accelerator chips growing?
Demand for AI accelerator chips is expected to grow 50% to 70% by 2028. 83% of buyers say they’ve already experienced supply issues—and new industrial uses will further increase demand.

Why are chip buyers in a bind?
Only a few companies can design and fabricate the latest high-performance AI chips, including graphics processing units and highly integrated smartphone chips. Essential machinery, materials, and tools are also made by a small number of firms, fabs, and foundries. 75% of chip-buying executives say depending on a small pool of semiconductor manufacturing vendors is a major strategic disadvantage. Long lead times and export controls add further supply constraints.

What are the biggest challenges faced by semiconductor providers?
More than 80% of supplier executives point to geopolitical tensions, investment constraints, and deficiencies in tech skills and supply chain capabilities. For example, 81% of chip suppliers say investment needed for new chip technologies is out of reach.

Why is it essential to have geographically diverse sources of chip suppliers?
80% of executives say it's essential to have access to local AI chips, AI talent pools, and AI platforms. But the transition is not happening fast enough for buyers and suppliers in the global semiconductor industry. By accessing locally produced AI chips, near-at-hand AI talent pools, and accessible AI platforms, regional sourcing can help increase ecosystem-wide innovation and reduce reliance on global supply chains, which can be vulnerable to disruptions and tariffs.

How is the growth of AI training models impacting global energy consumption?
Global demand for electricity by data centers is projected to more than double by 2030. To reduce energy consumption, 82% of data center and high-performance computing manufacturers are developing specialized AI chips, such as GPUs and integrated circuits optimized for specific tasks. In addition, 82% of chip buyers are pursuing specialized chips for specific tasks to optimize energy usage.

84% of semiconductor executives say government incentives play a crucial role in their organization’s growth and competitiveness.

What can chip buyers do now to secure their supply of advanced AI chips?

• Seek out regional partners with shared AI ambitions. 
  Identify diverse funding sources to support the development of regional semiconductor supply chains, including government grants, industry investments, and private sector partnerships.
• Realign sourcing, market, and partnership strategies to new global realities. Improve or acquire your own global intelligence capabilities. Prepare for unexpected business disruptions by strategically reallocating investments and procurement. Use AI to make scenario planning more accurate and comprehensive.
• Capture local AI demand with competitive AI computing capabilities. 
  Develop compelling, strong AI driven business cases supported by vertically integrated AI capabilities and proprietary data. Tap new AI revenue pools to nurture local AI ecosystems and semiconductor manufacturing capacity.
• Find out where customization can deliver a competitive advantage. 
  Assess AI driven capabilities that benefit most from custom components. Weigh those capabilities in the context of whether they will deliver differentiating outcomes. Pursue alternative technology approaches with strategic partnerships, such as advanced packaging chiplets, to gain AI competitiveness.
• Test the waters early. 
  Identify high-potential use cases for emerging technologies and volunteer to test or provide feedback on solutions as they evolve. Identify innovative ways to collaborate, such as sharing simulation environments or co-investing in pooled testing facilities, to share costs, reduce risk, and accelerate large-scale learning.

 
Download the report to learn how creativity, collaboration, and planning can compensate for AI chip shortages and help chipmakers and semiconductor buyers compete in the AI race.