the gdse framework a meta-tool for automated design space exploration tripti saxena graduate student...

26
The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

Upload: derick-jefferson

Post on 14-Dec-2015

215 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

1

The GDSE FrameworkA Meta-Tool for Automated Design Space Exploration

Tripti SaxenaGraduate Student

Vanderbilt University

Page 2: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

2

Outline

• Background• Motivation• The Generic Design Space Exploration

Framework– Reconfigurable Representation– Flexible Exploration

• Conclusion and Future Work

Page 3: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

3

Background

x2

x1designspace

Design Spaceproduct of possible discrete design choicese.g. selection of software components

alternative hardware architectures

selection of features

Design Space Explorationfind a design point • Satisfies constraints• Is “best” w.r.t. an

objective function(s)

Page 4: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

4

Background

• Software Product-line EngineeringFace Recognition System

Camera

Image CompressionPCA

MahCosineEuclidean

Bayesian

MLMAP

LDA

IdaSoftEuclidean

Face Recognition Algorithm

[1..4]

Feature Model

Page 5: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

5

Background

• Embedded Systems

TDMA

Priority

EDF

WFQ

RISC

DSP

LookUp

Cipher

Application

Mapping

Hardware

Page 6: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

6

Exploration Solver

Exploration algorithm

Exploration Solver

Exploration algorithm

Current DSE Frameworks

- Configured to solve a particular DSE problem- Supports only ONE solver/solving technique, not

efficient for every problem instance.

Representation Design space Constraints Objectives

Examples of DSE frameworks• FAMA• Milan• SPLOT• PISA• …

Page 7: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

Motivation

Hardware-Software Mapping

Software Product Line configuration

Web Server configuration

SAT(e.g. Minisat)

Mathematical Solver(e.g. LPSolve)

Constraint Solver(e.g Gecode)

Common core

Reconfigurable Representation Multiple Solvers

A reusable and flexible framework

7

Reusable Core

Page 8: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

The Generic Design Space Exploration Framework

DSMLADSEL

Template

eDSML

Design Space Model

Instance of

FlatZinc Solver

Solver Independent Constraint Problem in

Minizinc

Intermediate Language

Intermediate Design Space Model

Instance of

FD Solver

LP Solver

GecodeSolver

DESERT

Model Transformation

t

GME

8

Common Core

Reconfigurable Representation

Minizinc

Flexible Exploration

Page 9: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

9

Generic Modeling EnvironmentMetamodeling

Modeling

MDE-based Framework• Meta-Programmable• Reuse of previously

defined entities using libraries

• Translators for synthesis

Done by Domain experts

Done by Domain-engineers

Enables reconfigurable representation

Page 10: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

10

MinizincMedium Level

Solver-Independent Language

•Express Combinatorial Search Problems•Predefined translators for translation to different solver specific formats

InterpreterModel

Simplified Constraints + variables

Constraint Solver (Flatzinc)

Solutions

LP Solve Minisat

Solutions Solutions

Enables flexible exploration

Page 11: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

Overview of the GDSE Framework

DSML ADSEL

eDSML

Design Space Model

Instance of

FlatZinc Solver

Solver Independent Constraint Problem in

Minizinc

Intermediate Language

Intermediate Design Space Model

Instance of

FD Solver

LP Solver

GecodeSolver

DESERT

4

1

Model Transformation

t

2

3

GME

11

Reuse existing DSML

Page 12: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

12

Step 1: Domain Specific Modeling Language.

Metamodel•Entities•Relationships•Attributes

Metamodel

Face Recognition Algorithm

Model

Page 13: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

13

DSE Problem: Face Recognition System

DSML has to be extended to capture• Design Space of possible variants• DSE Properties • Memory• CPU• Cost

• Constraints• Bound constraint : Memory

<= x• ObjectivePCA

1. Construct a Face Recognition System

2. Goal: Choose a face recognition algorithm from the variants satisfying selection + resource constraints

Algo1

Algo2

Algo3…

Page 14: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

14

Step 2: Metamodel Composition

DSML ADSEL

eDSML

Design Space Model

Instance of

GME

FlatZinc Solver

Solver Independent Constraint Problem in

Minizinc

Intermediate Language

Intermediate Design Space Model

Instance of

FD Solver

LP Solver

GecodeSolver

DESERT

4

1

Model Transformation

t

2

3

GMEPerformed by

Domain-expert ONCE for a kind of DSE problem

Page 15: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

15

The Abstract Design Space Exploration Language Template

Objective

Design Space Tree

COMPONENT TYPES

CONSTRAINT TYPES OBJECTIVE TYPES

All elements are abstract !

Page 16: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

16

The ADSEL Component types

•PropertyType•ValueType•Domain•Composition Function

Page 17: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

17

The ADSEL Constraint and Objective types

e.g. A.Sel -> not B.Sel

e.g. A.Memory <= 100 e.g.minimize (cost)

e.g. utilization

Page 18: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

18

Metamodel Composition: Template Instantiation

Page 19: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

19

Metamodel Composition: Template Instantiation

Page 20: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

20

Composition Automation: eDSML Creator

• GUI• Semi-Automated Metamodel Creation based on user selection•Written in C++

Page 21: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

21

Step 3: Create Instance Model

NANA

M

Page 22: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

22

Step 4: Perform DSE

DSML ADSEL

eDSML

Design Space Model

Instance of

GME

FlatZinc Solver

Solver Independent Constraint Problem in

Minizinc

Intermediate Language

Intermediate Design Space Model

Instance of

FD Solver

LP Solver

GecodeSolver

DESERT

4

1

Model Transformation

t

2

3

GME

Page 23: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

23

Solutions

Solver Selection

Solver

Page 24: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

24

Summary

– A Generic Framework• Reusable• Flexible

– Case studies from different domains• Software Product Line Configuration • Architecture Synthesis • Hardware Software Co-synthesis

– Scalability: SPLE

Page 25: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

25

Conclusion

• Other use cases– Hybrid Tool: Invoke multiple solvers in series– Scalability Analysis Tool

• Future Work– Wider range of case studies– Support parametric representation– Support algorithms for multi-objective

optimization

Page 26: The GDSE Framework A Meta-Tool for Automated Design Space Exploration Tripti Saxena Graduate Student Vanderbilt University 1

26

Questions ?