Key Findings
At the 2026 Design Automation Conference (DAC), a critical session titled ‘The True Challenges of AI Chip Creation’ highlighted the extremely high complexity of Artificial Intelligence (AI) chip design and manufacturing processes, along with their dramatically escalating costs. Participating experts emphasized that addressing these challenges is essential for continuing innovation in AI chips.
Technical / Clinical Details
AI chips, particularly accelerators for large language models (LLMs) and generative AI, house billions of transistors and demand exceptionally high parallel processing capabilities and data bandwidth. Consequently, design complexity grows exponentially, requiring immense resources and time at each stage of physical design, circuit simulation, and verification. For example, the design cycle for a single cutting-edge AI chip is reported to be extended by several months to years compared to traditional general-purpose CPUs, necessitating hundreds of person-months of engineering resources. On the manufacturing front, the transition from FinFET to Gate-All-Around (GAA) transistors and the adoption of advanced packaging technologies (CoWoS, 3D stacking) further complicate processes, significantly increasing cost per wafer. Moreover, AI chips must be optimized for diverse AI workloads (training, inference, edge AI, etc.), requiring designs that balance flexibility and efficiency. The session presented solutions such as modularization through chiplet technology, leveraging open-source hardware IPs like RISC-V, and improving design efficiency by incorporating AI itself into Electronic Design Automation (EDA) tools.
Background & Context
The explosive growth of AI has spurred unprecedented investment and innovation in the semiconductor industry. However, with the evident slowdown of Moore’s Law, improvements in chip performance cannot be achieved without pushing the boundaries of design technology, materials science, and manufacturing processes. AI chips, in particular, constantly demand cutting-edge hardware to keep pace with software advancements. In this context, automation of design tools, accuracy of verification, and optimization of manufacturing costs have become bottlenecks in AI chip development. DAC serves as a primary forum for discussing frontier research and industry challenges in this domain.
Strategic Significance & Outlook
The challenges and solutions discussed at DAC 2026 will significantly influence the future direction of AI chip development. The further maturation of chiplet technology could enable modular designs, distributing design costs and risks. Furthermore, the introduction of AI-powered EDA tools is expected to accelerate design cycles and optimization, shortening the time-to-market for AI chips. It is crucial for investors to understand these technological trends and identify investment opportunities in key players within the AI chip design and manufacturing ecosystem (EDA vendors, foundries, IP providers). In the long term, these technological advancements are expected to drive down AI chip prices and promote their widespread adoption, laying the groundwork for AI technology to expand into a broader range of industries and applications.
Source: https://www.eetimes.com/dac-2026-what-does-it-actually-take-to-create-ai-chips/
Get our weekly technology intelligence — free
Receive an infographic that lets you judge at a glance whether each field’s analysis report is worth reading.
Subscribe Free — Weekly Tech Intelligence
By subscribing, you’ll receive Troy-Technical’s weekly technology intelligence newsletter.
- Your email and selected fields are used only to deliver the newsletter.
- We never share your information with third parties.
- You can unsubscribe anytime via the link in each email.
See our Privacy Policy for details.
Takes about a minute · Unsubscribe anytime
Comments