A naukri.com initiative
Semiengineering
5d
380
Image Credit: Semiengineering
First-Time Silicon Success Plummets
- First-time silicon success is declining sharply due to increased complexity, more iterations, and customization in chip design, making chips less stable and requiring more optimization and iterations.
- The rise in logic, SRAM, interconnects, and software, along with the need for advanced cooling techniques, new materials, and more verification, contribute to the challenges faced by chipmakers.
- Historically, first-silicon success rates were around 30%, dropping to 14% in recent times, with the majority of projects falling behind schedule, reflecting the industry's lowest success point.
- The productivity gap necessitates increased productivity, but talent shortages and training constraints hinder this, prompting EDA vendors to integrate AI into their tools to enhance productivity.
- Industry shifts towards more specialized chips have accelerated chip production, leading to a surge in tape-outs but without a proportional increase in staffing, exacerbating productivity challenges.
- Design errors, design changes, and misunderstandings of specifications contribute to respins and complexity issues, highlighting the need for improved design processes and tools.
- Incorporating AI into EDA tools offers generative and agentic AI capabilities, streamlining design processes, improving productivity, and enhancing collaboration between engineers and AI systems.
- Increased software integration, complexity, and co-design requirements pose challenges in chip verification, necessitating a shift in workflow and balance between hardware and software functionalities.
- Safety-critical markets like automotive demand reliable, high-performance chips, driving the need for advanced technologies and customized packages that prioritize safety and security.
- Addressing security vulnerabilities, design changes, and software-hardware co-design complexities require enhanced verification methodologies and testing procedures to achieve first-time silicon success.
- Overall, the industry faces a pressing need to adapt to growing complexity, talent shortages, and productivity challenges by leveraging AI, improving design workflows, and enhancing collaboration among engineers.
Read Full Article
22 Likes
For uninterrupted reading, download the app