Walking Pads Framework for Chip Pad Placement Optimization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for optimizing power pad placement in electric devices are inefficient in addressing transient voltage noise and temperature violations, due to high computational complexity and time-consuming simulations, which hinder effective transient noise mitigation in modern SoC design.

Innovation Solution

The Walking Pads (WP) framework converts the global pad placement optimization problem into a local virtual-force balance problem, using a virtual-force-directed pad-movement strategy and statistical sampling to reduce computation cost and optimize power supply pad placement for transient noise control, incorporating practical optimization (PO) and deep optimization (DO) algorithms for varying requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If steady-state IR-drop based pad placement optimization is used, then the manufacturing process is simple and computational cost is low, but transient voltage noise violations are not effectively mitigated

Engineering Contradiction:
Improvetransient voltage noise controlVSAvoidcomputational complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent replaces complex transient voltage simulation with a simplified IR-drop based optimization model. Instead of performing full transient simulations to evaluate pad placement, the system uses IR-drop calculations which are computationally much less intensive, while still achieving effective transient noise mitigation through the virtual force directed movement strategy

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the evaluation parameter from full transient voltage analysis to IR-drop metrics. By using IR-drop as a proxy for transient noise evaluation, the system achieves comparable optimization results with significantly reduced computational complexity, enabling practical implementation in industrial design flows

Inventive Principle:
Principle #35Parameter changes

2Manufacturing precision

If exhaustive search methods are used for pad placement optimization, then manufacturing precision is high, but productivity is low due to enormous design space

Engineering Contradiction:
Improvepad placement optimization qualityVSAvoidoptimization speed
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The patent implements a self-service optimization mechanism where pads automatically move to optimal positions based on virtual force calculations. The system evaluates all candidate pad locations simultaneously and directs movement without requiring exhaustive iterative searches, achieving both high optimization quality and fast convergence in a single evaluation pass

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent performs preliminary evaluation of all pad placement candidates using IR-drop calculations before final optimization. By pre-computing the virtual forces and optimal movement directions for all pads, the system avoids time-consuming iterative searches and directly achieves the optimal configuration in one step

Inventive Principle:
Principle #10Preliminary action

3Reliability

If more C4 pads are allocated for power delivery, then voltage noise is reduced, but I/O bandwidth is reduced

Engineering Contradiction:
Improvevoltage supply stabilityVSAvoidI/O bandwidth
Core Design Contradiction:
ReliabilityVSSpeed

Solution Approach 1:

The patent applies local quality optimization by strategically placing power pads in specific locations where they are most effective for noise mitigation. Instead of uniformly distributing power pads across the chip, the system identifies critical regions with high transient noise risk and concentrates pad placement in those areas, achieving effective noise control with fewer total pads

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent uses virtual force models to create a simplified representation of the complex transient noise problem. By copying the essential characteristics of transient noise behavior into an IR-drop based virtual force model, the system can optimize pad placement without requiring actual transient simulations, enabling efficient exploration of the design space to find the optimal trade-off between power delivery and I/O bandwidth

Inventive Principle:
Principle #26Copying

Data Source

PatentUS11436401B2System for placement optimization of chip design for transient noise control and related methods thereof
Publication Date: 2022.09.06 UNIV OF VIRGINIA
  • US11436401B2 patent drawing
  • US11436401B2 patent drawing
  • US11436401B2 patent drawing

AI summary

Transient voltage noise, including resistive and reactive noise, causes timing errors at runtime. A heuristic framework, Walking Pads, is introduced to minimize transient voltage violations by optimizing power supply pad placement. It is shown that the steady-state optimal design point differs from the transient optimum, and further noise reduction can be achieved with transient optimization. The methodology significantly reduces voltage violations by balancing the average transient voltage noise of the four branches at each pad site. When pad placement is optimized using a representative stressmark, voltage violations are reduced 46-80% across 11 Parsec benchmarks with respect to the results from IR-drop-optimized pad placement. It is shown that the allocation of on-chip decoupling capacitance significantly influences the optimal locations of pads.