Abstract:The increasing complexity of digital designs and the requirement of timing measurements in various design stages make static timing analysis critical. Each design stage utilizes static timing analysis to evaluate the system performance, and then optimizes the design accordingly. An accurate and efficient timing analysis package is crucial for the success of the whole design process. We studied three important problems in static timing analysis: false paths, multi-cycle paths, and hierarchical timing analysis. Static timing analysis should deal with false paths and multi-cycle paths to produce accurate timings. Previous published works focused on dealing with false paths in static timing analysis. There are several approaches; such as labeling algorithm and node-splitting approach, using tags to label and eliminate false path timings. A large number of tags need to be created and propagated. Thus, the efficiency is deteriorated. For hierarchical timing analysis, timing analysis iteratively performed on flatten circuits suffers from low efficiency because of increasing design complexity. Inspired by the gap between challenges and current sub-optimal solutions, we proposed three new techniques: (1) A two-direction propagation is proposed which minimizes the number of tagged false path timings using a biclique covering approach. We proved that all the non-false paths are covered and all the false paths are removed. A polynomial heuristic to perform the biclique covering minimization in minimal degree order is also introduced. (2) A framework unified processing false paths and multi-cycle paths is proposed which expands the tag-based approach for false paths to cover multi-cycle paths. Following the biclique covering approach for false paths, we devise time shifting for multi-cycle paths to improve the efficiency. We prove that the unified framework produces accurate timings. (3) An abstract, timing model reduction technique is proposed for hierarchical timing analysis. We introduce an iterative biclique-star replacement technique to minimize the abstract timing model, thus improving the efficiency. Combined the techniques proposed above, we provide an accurate and efficient static timing analysis package, which can be used in hierarchical design methodology.

A Fast and Simple Block-Based Approach for Common Path Pessimism Removal in Static Timing Analysis

A Fast And Accurate Approach For Common Path Pessimism Removal In Static Timing Analysis

A Provably Good and Practically Efficient Algorithm for Common Path Pessimism Removal in Large Designs

Heterogeneous Static Timing Analysis with Advanced Delay Calculator

Dynamic Supply Noise Aware Timing Analysis with JIT Machine Learning Integration

HeteroCPPR: Accelerating Common Path Pessimism Removal with Heterogeneous CPU-GPU Parallelism

Efficient Static Timing Analysis Using a Unified Framework for False Paths and Multi-Cycle Paths

Design of non-critical path resource distributor for SMT processors

GPU-accelerated Critical Path Generation with Path Constraints

A Dedicated Greedy Short Path Padding Solution Method For Error Resilient Circuit Designs

Static timing analysis in vlsi design

Improving the Efficiency of Static Timing Analysis with False Paths

TimFastPlace: Critical-path Based Timing Driven FastPlace

A General Graph Based Pessimism Reduction Framework for Design Optimization of Timing Closure.

GPU-accelerated Path-based Timing Analysis.

Eliminating Timing Errors Through Collaborative Design to Maximize the Throughput

A New Approach to Solving False Path Problem in Timing Analysis.

Fast Statistical Timing Analysis for Circuits with Post-Silicon Tunable Clock Buffers

Adaptive Stochastic Collocation Method For Parameterized Statistical Timing Analysis With Quadratic Delay Model

A scan chain optimization method for diagnosis

Efficient Timing Analysis with Known False Paths Using Biclique Covering