Abstract:To supportincremental changeson theexisting floorplan derived fromsomechanges inthehighlevel synthesis isbecoming anincreasingly important problem inordertocope withthecomplexity ofthemerging ofVLSIdesign phases and thenecessary ofalternation betweensynthesis andphysical design. Inthis paper, wepresent anincremental floorplanning algorithm withLessFlexibility First (LFF)Principles (LFF_IFP algorithm). LFF principle isadeterministic algorithm through human'saccumulated experience. Inthispaper, thehorizontal andvertical constraint graphs arebuilt upbasedonthepacking oftheinitial floorplanning result within linear time, andthe accumulated slackdistances amongblocks arecomputed. The flexibility ofblocks tobe inserted totheinitial floorplan ('incremental block' isalsocalled hereforshort) iscomputed and every slack spacebelongs totheblockintheinitial floorplan also hasdifferent degree offlexibility foragiven incremental block. Theexperimental results demonstrate thatthealgorithm is effective andpromising inbuilding blocklayout application. I.INTRODUCTION Inthepresent, everydesign phasehasdifferent design hierarchy andoptimization goalwhichleads todifferent netlist information andblock size. Otherwise, itoften happens that thelayout needs locally modification after floorplanning inorder tosatisfy therequirement ofdelay, powerorrouting completion rate, orimprove thecircuit performance. Inthis case, theinitial floorplan configuration mustbemodified ina quick timetoanewfloorplan whichcanalsomaintain the original goodperformance orevenbetter. Thisisincremental floorplanning problem (IFP). IFPcanberegarded asan iterative amendment process. Withthecontinuous shrinking offeature size, deepsubmicron effects causemoreandmore interactions between high-level andphysical-level design and itisveryimportant forthedesign convergence. Suchas different binding, allocation strategies inthehighlevel synthesis canlead tothechange oftheblocks' size orshape oramountinthephysical design. Thisisalso thereason why IFPexists. Recently, several synthesis researches havebeen reported taking physical information intoaccount, suchas

VLSI Placement with Arbitrary Rectilinear Block with Less Flexibility First Principles

VLSI Block Placement Using Less Flexibility First Principles

Deterministic VLSI block placement algorithm using Less Flexibility First principle

Arbitrary Rectilinear Block Packing Based on Less Flexibility First Principles

Module Placement On Arbitrarily Rectilinear Regions Using Less Flexibility First Principles

VLSI Placement with Pre-Placed Modules Based on Less Flexibility First Principles

Interconnection driven placement algorithm based on LFF principle

An Incremental Floorplanning Algorithm with Less Flexibility First Principles

Interconnection Driven VLSI Module Placement Based on Quadratic Programming and Considering Congestion Using LFF Principles

VLSI Floorplan Based on Less Flexibility First Principle and Linear Programming

A New Approach Based on Lff for Optimization of Dynamic Hardware Reconfigurations

A Comparison of Less Flexibility First Principles with Simulated Annealing

A FastPlacement Approach Forlargescale Modules Basedon LessFlexibility First Principles

General Floorplans with L/T-shaped Blocks Using Corner Block List

Large Scale VLSI Module Placement Using LFF Heuristics by Stages

Vlsi Floorplanning with Boundary Constraints Using Corner Block List Representation

Module Placement Based On Quadratic Programming And Rectangle Packing Using Less Flexibility First Principle

DREAMPlace 3.0: Multi-Electrostatics Based Robust VLSI Placement with Region Constraints

VLSI Floorplanning with Boundary Constraints Based on Corner Block List

A New Timing-Driven Placement Algorithm Based on Table-Lookup Delay Model

AnIncremental FloorplanningAlgorithm with LessFlexibility First Principles