SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto...
-
date post
19-Dec-2015 -
Category
Documents
-
view
221 -
download
1
Transcript of SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto...
![Page 1: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/1.jpg)
SDPA:Leading-edge Software for SDP
2008/10/14 @ Informs ’08
Tokyo Institute of Technology Makoto YamashitaMituhiro FukudaMasakazu KojimaKazuhide Nakata
Chuo University Katsuki Fujisawa
National Maritime Research Institute Kazuhiro Kobayashi
RIKEN Maho Nakata
![Page 2: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/2.jpg)
2
SDPA (SemiDefinite Programming Algorithm) Project
Open Source Software to solveSemiDefinite Programming
Since the 1st release in 1995, it has kept high quality
In 2008, the latest version SDPA 7 was released and has been updated continuously
Many software for more advantage
![Page 3: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/3.jpg)
3
SDPA Family
SDPA
SDPARA(Parallel with MPI)
SDPA-C(Matrix Completion)
SDPA-M(Matlab Interface)
SDPA-GMP(Multiple Precision)
SDPARA-C
accessible onSDPA Online Solveras Web service
![Page 4: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/4.jpg)
4
Outline of this talk
1. SDP and the improvements of SDPA72. Parallel with MPI3. Multiple Precision4. Online Solver5. Future Works
![Page 5: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/5.jpg)
6
Standard form of SDP
![Page 6: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/6.jpg)
7
Applications of SDP
Control Theory Lyapnov condition
Combinatorial Optimization Max Cut Theta function
Quantum Chemistry Reduced Density Matrix
![Page 7: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/7.jpg)
8
Primal-Dual Interior-Point Methods
![Page 8: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/8.jpg)
9
Computation for Search Direction
Schur complement matrix ⇒ Cholesky Factorizaiton
Exploitation of Sparsity in
![Page 9: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/9.jpg)
11
SDPA 7
SDPA7 resolves bottlenecks of SDPA6 Introduce sparse Cholesky factorization
for the Schur complement matrix Adopt new data structure Reduce memory space for temporary
variables Introduce configure script for easier
installation
![Page 10: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/10.jpg)
12
Sparsity pattern of Schur complement matrix
Fully dense Schur complement matrixFully dense Schur complement matrixSparse Schur complement matrixSparse Schur complement matrixminimum degree ordering minimum degree ordering to minimize to minimize the number of fill-in the number of fill-in
![Page 11: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/11.jpg)
13
New Data Structure For multiple diagonal
structure SDPA6
stores nonzero elements of each block
stores all blocks SDPA 7
stores nonzero elements of each block
stores only nonzero blocks
![Page 12: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/12.jpg)
14
r2S_broydenTri300.dat-s
SDPA 6 SDPA 7
Computing B 1009.0s 0.8s
Cholesky B 5179.5s 0.5s
Total CPU Time 6204.4s 2.8s
Input data 90MB 1MB
Matrix B 272MB 4MB
Dense Matrix 7MB 3MB
Total Memory 380MB 19MB
Sparse Schur
New Data Structure
Efficient Temporary
Xeon 2.80GHz, 2GB memory, Linux 2.4
![Page 13: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/13.jpg)
15
Configure script
Easier installation$./configure –with-blas=”-lblas”
–with-lapack=”-llapack”$ make$ make install
We can link with Optimized BLAS, i.e., ATLAS, GotoBLAS, Intel MKL
![Page 14: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/14.jpg)
16
Matlab Interface
SDPA-M is the Matlab interface [mDIM,nBLOCK,bLOCKsTRUCT,c,F] =
read_data(’example1.dat-s’); [objVal,x,X,Y,INFO] =
sdpam(mDIM,nBLOCK,bLOCKsTRUCT,c,F,OPTION);
SeDuMi Input interface [At,b,c,K] = fromsdpa(’example1.dat-s’); [x,y,info] = sedumiwrap(At,b,c,K,[],pars);
Current version is for only LP and SDP cones Parameter control is based on SDPA
![Page 15: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/15.jpg)
17
Extremely Large Problem and Bottlenecks
The largest size requires 8.6GB memory
We replace these two bottlenecks by parallel computation
Opteron 246 2.0GHz6GB Memory
![Page 16: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/16.jpg)
18
Exploitation of Sparsityin SDPA
We change the formula by row-wise
We keep this scheme on parallel computation
F1
F2
F3
![Page 17: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/17.jpg)
19
Row-wise distribution for evaluation of the Schur complement matrix
4 CPU is availableEach CPU computes only their assigned rows
. No communication between CPUsEfficient memory management
![Page 18: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/18.jpg)
20
Parallel Cholesky factorization We adopt Scalapack for the Cholesky
factorization of the Schur complement matrix We redistribute the matrix from row-wise to
two-dimensional block-cyclic-distribtuion
Redistribution
![Page 19: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/19.jpg)
21
Computation Time for NH3
93883235
1077460
9150
2973
796
201
88
239.4
3.71
10
100
1000
10000
1 4 16 64#processors
seco
nd TOTALComp BChol B
TSUBAME@Tokyo-TechOpteron 880 (2.4GHz)32GB memory/node
![Page 20: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/20.jpg)
22
Scalability for LiF
1
10
100
1 4 16 64#processors
scal
abili
ty TOTALComp BChol B
Total 28 times
Comp B 43 times
Chol B 46 times
Row-wise distribution for Comp B is very effective
TSUBAME@Tokyo-TechOpteron 880 (2.4GHz)32GB memory/node
![Page 21: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/21.jpg)
26
Multiple Precision SDPA uses ‘double’ precision
53 significant bit almost 8 digit
SDPA 7 result (gpp124-1 from SDPLIB) Objective Function (Only 5 digits)
-7.3430761748645921e+00(Primal)-7.3430800814821620e+00(Dual)
Feasibility5.45696821063e-12 (Primal)1.68252292320e-07 (Dual)
Some applications requires more accuracy
![Page 22: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/22.jpg)
27
SDPA-GMP
GMP: Gnu Multiple Precision Library Arbitrary fixed precision ‘double’ precision is replaced by GMP Ultra High Accuracy
by long computation time
![Page 23: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/23.jpg)
28
Ultra Accuracy of SDPA-GMP
![Page 24: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/24.jpg)
29
Comparison on SDPA and SDPA-GMP(384bit)
gpp124-1(SDPLIB)
SDPA-GMP(7.1.0)Relative gap 1.7163710368162993e-26Objective Function -7.3430762652465377e+00 (Primal) -7.3430762652465377e+00 (Dual)Feasibility 2.0710194844721e-57 (Primal) 1.2329417039702e-29 (Dual)Computation time 228.95 sec. (59 iterations)
SDPA(7.1.0)Relative gap 5.3201361904260111e-07Objective Function -7.3430761748645921e+00 (Primal) -7.3430800814821620e+00 (Dual)Feasibility 5.45696821063e-12 (Primal) 1.68252292320e-07 (Dual)Computation time 0.14 sec. (20 iterations)
![Page 25: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/25.jpg)
30
SDPA Online Solver
SDPA Online Solver will offer SDPA/SDPARA/SDPARA-C via the Internet.
Internet
InterfaceUser1.Input 2.Ninf-G
3.SDPARA on PC cluster
4.Solution
![Page 26: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/26.jpg)
31
To use Online Solver Users without parallel environment can
use SDPARA/SDPARA-C. No Charge. Registration via the Internet is required
so that passwords to protect users data will be generated automatically.
Access SDPA Project Home Page.[SDPA Online for your future.]http://sdpa.indsys.chuo-u.ac.jp/sdpa/
![Page 27: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/27.jpg)
32
Online Solver Interface
![Page 28: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/28.jpg)
33
Online Solver Usage
![Page 29: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/29.jpg)
34
Conclusion The latest version 7 attains higher
performance than version 6 Parallel Solver enables us to solve
extremely large SDPs Matrix Completion is useful for Structural
Sparsity SDPA-GMP generates ultra high accuracy
solution Online Solver provides powerful
computation resources via the Internet
![Page 30: SDPA: Leading-edge Software for SDP 2008/10/14 @ Informs ’ 08 Tokyo Institute of Technology Makoto Yamashita Mituhiro Fukuda Masakazu Kojima Kazuhide Nakata.](https://reader036.fdocuments.net/reader036/viewer/2022062421/56649d2c5503460f94a01aaf/html5/thumbnails/30.jpg)
35
Future works
Callable Library of SDPA7 Automatic Selection from
SDPA/SDPARA/SDPARA-C