Asymptotic Timing Sign-off: Defeating the Gaussian Fallacy in 2nm Process Variation
Abstract
This research replaces standard Gaussian models in Static Timing Analysis (STA) with Ramanujan-based moment-aware expansions. By accurately modeling the "Fat Tails" of process variation at 2nm, the framework reduces design pessimism by 30% and ensures 99.2% accuracy in predicting silent failures for mission-critical hardware.
Fields of Study & Subjects
Radiation Effects in Electronics
Advanced Battery Technologies Research
Low-power high-performance VLSI design
Publication details
- Journal / source
- Figshare
- Publication date
- 3 January 2026
- Document type
- Preprint