Power Grid Sizing via Convex Programming Peng Du, Shih-Hung Weng, Xiang Hu, Chung-Kuan Cheng University of California, San Diego 1.

Slides:



Advertisements
Similar presentations
THERMAL-AWARE BUS-DRIVEN FLOORPLANNING PO-HSUN WU & TSUNG-YI HO Department of Computer Science and Information Engineering, National Cheng Kung University.
Advertisements

ECE Longest Path dual 1 ECE 665 Spring 2005 ECE 665 Spring 2005 Computer Algorithms with Applications to VLSI CAD Linear Programming Duality – Longest.
Exploring 3D Power Distribution Network Physics
A Graph-Partitioning-Based Approach for Multi-Layer Constrained Via Minimization Yih-Chih Chou and Youn-Long Lin Department of Computer Science, Tsing.
1 Advancing Supercomputer Performance Through Interconnection Topology Synthesis Yi Zhu, Michael Taylor, Scott B. Baden and Chung-Kuan Cheng Department.
Paul Falkenstern and Yuan Xie Yao-Wen Chang Yu Wang Three-Dimensional Integrated Circuits (3D IC) Floorplan and Power/Ground Network Co-synthesis ASPDAC’10.
Presented by: Hao Liang
Graph Laplacian Regularization for Large-Scale Semidefinite Programming Kilian Weinberger et al. NIPS 2006 presented by Aggeliki Tsoli.
Efficient Escape Routing Rui Shi, Chung-Kuan Cheng University of California, San Diego.
1 Thermal Via Placement in 3D ICs Brent Goplen, Sachin Sapatnekar Department of Electrical and Computer Engineering University of Minnesota.
Visual Recognition Tutorial
Path Finding for 3D Power Distribution Networks A. B. Kahng and C. K. Cheng UC San Diego Feb 18, 2011.
A Beacon-Less Location Discovery Scheme for Wireless Sensor Networks Lei Fang (Syracuse) Wenliang (Kevin) Du (Syracuse) Peng Ning (North Carolina State)
Layer Assignment Algorithm for RLC Crosstalk Minimization Bin Liu, Yici Cai, Qiang Zhou, Xianlong Hong Tsinghua University.
Adaptive Data Collection Strategies for Lifetime-Constrained Wireless Sensor Networks Xueyan Tang Jianliang Xu Sch. of Comput. Eng., Nanyang Technol. Univ.,
On the Construction of Energy- Efficient Broadcast Tree with Hitch-hiking in Wireless Networks Source: 2004 International Performance Computing and Communications.
Supply Voltage Degradation Aware Analytical Placement Andrew B. Kahng, Bao Liu and Qinke Wang UCSD CSE Department {abk, bliu,
Chung-Kuan Cheng†, Andrew B. Kahng†‡,
Visual Recognition Tutorial
1 Traffic Shaping to Optimize Ad Delivery Deepayan Chakrabarti Erik Vee.
Javad Lavaei Department of Electrical Engineering Columbia University Low-Rank Solution for Nonlinear Optimization over Graphs.
More Realistic Power Grid Verification Based on Hierarchical Current and Power constraints 2 Chung-Kuan Cheng, 2 Peng Du, 2 Andrew B. Kahng, 1 Grantham.
Introduction to Optimization (Part 1)
Javad Lavaei Department of Electrical Engineering Columbia University Joint work with Somayeh Sojoudi Convexification of Optimal Power Flow Problem by.
Advanced Image Processing Image Relaxation – Restoration and Feature Extraction 02/02/10.
Dose Map and Placement Co-Optimization for Timing Yield Enhancement and Leakage Power Reduction Kwangok Jeong, Andrew B. Kahng, Chul-Hong Park, Hailong.
University of California San Diego
A Topology-based ECO Routing Methodology for Mask Cost Minimization Po-Hsun Wu, Shang-Ya Bai, and Tsung-Yi Ho Department of Computer Science and Information.
Authors: Jia-Wei Fang,Chin-Hsiung Hsu,and Yao-Wen Chang DAC 2007 speaker: sheng yi An Integer Linear Programming Based Routing Algorithm for Flip-Chip.
Research on Analysis and Physical Synthesis Chung-Kuan Cheng CSE Department UC San Diego
Pattern Selection based co-design of Floorplan and Power/Ground Network with Wiring Resource Optimization L. Li, Y. Ma, N. Xu, Y. Wang and X. Hong WuHan.
Minimum Phoneme Error Based Heteroscedastic Linear Discriminant Analysis for Speech Recognition Bing Zhang and Spyros Matsoukas BBN Technologies Present.
1 中華大學資訊工程學系 Ching-Hsien Hsu ( 許慶賢 ) Localization and Scheduling Techniques for Optimizing Communications on Heterogeneous.
Low-Power Gated Bus Synthesis for 3D IC via Rectilinear Shortest-Path Steiner Graph Chung-Kuan Cheng, Peng Du, Andrew B. Kahng, and Shih-Hung Weng UC San.
UC San Diego / VLSI CAD Laboratory Incremental Multiple-Scan Chain Ordering for ECO Flip-Flop Insertion Andrew B. Kahng, Ilgweon Kang and Siddhartha Nath.
1 CS612 Algorithms for Electronic Design Automation CS 612 – Lecture 8 Lecture 8 Network Flow Based Modeling Mustafa Ozdal Computer Engineering Department,
Statistical Sampling-Based Parametric Analysis of Power Grids Dr. Peng Li Presented by Xueqian Zhao EE5970 Seminar.
Optimization Flow Control—I: Basic Algorithm and Convergence Present : Li-der.
9-1 Chapter 9 Project Scheduling Chapter 9 Project Scheduling McGraw-Hill/Irwin Copyright © 2005 by The McGraw-Hill Companies, Inc. All rights reserved.
Peng Du, Wenbo Zhao, Shih-Hung Weng, Chung-Kuan Cheng, Ronald Graham CSE Dept., University of California, San Diego, CA Character Design and Stamp Algorithms.
TSV-Constrained Micro- Channel Infrastructure Design for Cooling Stacked 3D-ICs Bing Shi and Ankur Srivastava, University of Maryland, College Park, MD,
1 Markov random field: A brief introduction (2) Tzu-Cheng Jen Institute of Electronics, NCTU
1 Iterative Integer Programming Formulation for Robust Resource Allocation in Dynamic Real-Time Systems Sethavidh Gertphol and Viktor K. Prasanna University.
Linear Models for Classification
Jianhua Liu1, Yi Zhu1, Haikun Zhu1, John Lillis2, Chung-Kuan Cheng1
Analytic Placement Algorithms Chung-Kuan Cheng CSE Department, UC San Diego, CA Contact: 1.
LEMAR: A Novel Length Matching Routing Algorithm for Analog and Mixed Signal Circuits H. Yao, Y. Cai and Q. Gao EDA Lab, Department of CS, Tsinghua University,
Speech Lab, ECE, State University of New York at Binghamton  Classification accuracies of neural network (left) and MXL (right) classifiers with various.
Efficient Resource Allocation for Wireless Multicast De-Nian Yang, Member, IEEE Ming-Syan Chen, Fellow, IEEE IEEE Transactions on Mobile Computing, April.
BME 353 – BIOMEDICAL MEASUREMENTS AND INSTRUMENTATION MEASUREMENT PRINCIPLES.
Circuit Simulation using Matrix Exponential Method Shih-Hung Weng, Quan Chen and Chung-Kuan Cheng CSE Department, UC San Diego, CA Contact:
Optimal Relay Placement for Indoor Sensor Networks Cuiyao Xue †, Yanmin Zhu †, Lei Ni †, Minglu Li †, Bo Li ‡ † Shanghai Jiao Tong University ‡ HK University.
Determining Optimal Processor Speeds for Periodic Real-Time Tasks with Different Power Characteristics H. Aydın, R. Melhem, D. Mossé, P.M. Alvarez University.
Character Design and Stamp Algorithms for Character Projection Electron-Beam Lithography P. Du, W. Zhao, S.H. Weng, C.K. Cheng, and R. Graham UC San Diego.
Techniques of Circuit Analysis
Exploring the Rogue Wave Phenomenon in 3D Power Distribution Networks Xiang Hu 1, Peng Du 2, Chung-Kuan Cheng 2 1 ECE Dept., 2 CSE Dept. University of.
Unified Adaptivity Optimization of Clock and Logic Signals Shiyan Hu and Jiang Hu Dept of Electrical and Computer Engineering Texas A&M University.
Puzzle You have 2 glass marbles Building with 100 floors
Chapter 3 Resistive Network Analysis Electrical Engineering/GIEE
Goal We present a hybrid optimization approach for solving global optimization problems, in particular automated parameter estimation models. The hybrid.
On-Chip Power Network Optimization with Decoupling Capacitors and Controlled-ESRs Wanping Zhang1,2, Ling Zhang2, Amirali Shayan2, Wenjian Yu3, Xiang Hu2,
Xiang Hu1, Wenbo Zhao2, Peng Du2, Amirali Shayan2, Chung-Kuan Cheng2
Haihua Su, Sani R. Nassif IBM ARL
Presented by Rich Goyette
Lecture 11. MLP (III): Back-Propagation
Neural Networks for Vertex Covering
Yiyu Shi*, Wei Yao*, Jinjun Xiong+ and Lei He*
3.3 Network-Centric Community Detection
Under a Concurrent and Hierarchical Scheme
Rong Ge, Duke University
Presentation transcript:

Power Grid Sizing via Convex Programming Peng Du, Shih-Hung Weng, Xiang Hu, Chung-Kuan Cheng University of California, San Diego 1

Outline Problem Formulation Convex Programming Reduction Optimizer of the Convex Programming Problem – Interior point and gradient descent methods – A Krylov space method to evaluate effective resistances – Close form of the derivative of effective resistances Experimental Results 2

Problem Formulation 3 Power Network N(V,E) Voltage source at node u Current loads in set W Variables: conductance g(e) for each edge e Objective: min max voltage drop at all nodes and over all possible current loads

Problem Formulation 4 D(v,g,I) : the voltage drop between nodes u and v for given conductance assignment g and current profile I. We assume the current profile satisfies the normalization constraint I(w 1 )+I(w 2 )+…+I(w r )=1 where W={w 1,w 2,…,w r }.

Convex Programming Reduction 5

SDP Formulation 6

Optimizer for Larger Cases Interior point method. Gradient descent method. Obstacles: – Evaluate the objective: effective resistances. – Evaluate the derivative of effective resistances relative to conductance. 7

Evaluate the Effective Resistances 8

Evaluate the Derivative of Effective Resistances 9

Experimental Results (Regular Grids) 10

Experimental Results (Regular Grids) The effective resistances before (blue plane) and after (green plane) optimization. 11

Experimental Results (Regular Grids) (R,C)MaxR(uni)MaxR(opt)Improvement (4,4) % (7,5) % (10,10) % 12

Experimental Results (Practical Power Grid) 13

Experimental Results (Practical Power Grid) 14

Voltage Map of Layer M1 15 The worst voltage drop is achieved near the origin which is the farthest point to the voltage source.

Resource Transformation between M6 and AP “Adjust Percentage” indicates the ratio of M6 resource obtained from AP. A monotonically decreasing curve means that with more resource to M6 we can reduce worst voltage drop. 16

Conclusion A power grid sizing method to minimize the maximum voltage drop over all test locations and current source distributions. We reduce the original problem into a convex programming problem whose objective is to minimize the maximum effective resistance. We adopt a Krylov space method to evaluate the effective resistances and devise a formula to update the derivative of effective resistance. Experimental results show that our method can achieve up to 40% improvement for regular 2D grids and 7.32% improvement for a practical power grid with only top two layers tunable.