A Modified Genetic Algorithm for the Selection of Decoupling Capacitors in PDN Design

Juang, Jack; Zhang, Ling; Kiguradze, Zurab; Pu, Bo; Jin, Shuai; Hwang, Chulsoon

doi:10.1109/emc/si/pi/emceurope52599.2021.9559292

Cited by 11 publications

(31 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Genetic Algorithm-based Methods. The most competitive non-learning method for DPP is the genetic algorithm (GA) [11,12,13,14]. GA is easy-to-implement in black-box optimization tasks.…”

Section: Related Workmentioning

confidence: 99%

“…Thus, most works achieved high performance in the devised problems only but did not quantitatively evaluated sample-efficiency. Most of the previous GA-based DPP works [11,13,14] have contributions regarding the objective function rather than the method itself. [12] proposed a GA method that finds the placement of decaps with minimum number that meets an arbitrary target impedance.…”

Section: A Extended Related Workmentioning

confidence: 99%

“…We used a genetic algorithm (GA) as the expert policy to collect expert guiding labels for imitation learning. GA is the most widely used search heuristic method for DPP [11,12,13,14]. We devised our own GA for DPP, the objective of which is to find the placement of given number (K) of decaps on PDN with a probing port and 0-15 keep-out regions that best suppresses the impedance of the probing port.…”

Section: Expert Label Collectionmentioning

confidence: 99%

“…Probability calculations using h (c) , and h i , i ∈ {1, ..., N × M } in (11)(12)(13)(14) are exactly identical to (5)(6)(7)(8) in [19] except the masking mechanism in equation 13 and equation 14. Because [19] solves TSP, so they mask the previously selected actions by forcing −∞ as compatibility u (c)j .…”

Section: D3 Calculation Of Probabilitymentioning

confidence: 99%

“…Thus, PI evaluation can not be formulated as a closed analytical form, making conventional mixed-integer linear programming (MILP)-based combinatorial optimization approaches not applicable. Therefore, classical genetic algorithm (GA)-based methods [11,12,13,14], black-box optimization methods, have long been used to solve DPP. Despite its effectiveness and efficient exploration over a huge design space, GA unavoidably requires many simulations and is not reusable; GA undergoes massive iterations for each given DPP task (i.e., sample inefficient).…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Collaborative Distillation Meta Learning for Simulation Intensive Hardware Design

Kim¹,

Kim²,

Kim³

et al. 2022

Preprint

View full text Add to dashboard Cite

This paper proposes a novel collaborative distillation meta learning (CDML) framework for simulation intensive hardware design problems. Deep reinforcement learning (DRL) has shown promising performance in various hardware design problems. However, previous works on DRL-based hardware design only dealt with problems with simplified objectives, which are not practical. In fact, the objective evaluation of real-world electrical performance through simulation is costly in terms of both time and computation, making DRL scheme involving extensive reward calculations not suitable. In this paper, we apply the CDML framework to decoupling capacitor placement problem (DPP), one of the significant simulation intensive hardware design problems. The CDML framework consists of a context-based meta learner and collaborative distillation scheme to produce a reusable solver. The context-based meta learner captures the location of probing port (i.e., target circuit block) and improves generalization capability. The collaborative distillation scheme with equivariant label transformation imposes the action-permutation (AP)-equivariant nature of placement problems, which not only improves sample efficiency but also improves generalization capability. Extensive experimental results verified that our CDML outperforms both neural baselines and iterative conventional design methods in terms of real-world objective, power integrity, with zero-shot transfer-ability.Preprint. Under review.

show abstract

“…Genetic Algorithm-based Methods. The most competitive non-learning method for DPP is the genetic algorithm (GA) [11,12,13,14]. GA is easy-to-implement in black-box optimization tasks.…”

Section: Related Workmentioning

confidence: 99%