Compiling and Using the

1
Compiling and Using the best R

• Vipin Sachdeva
• IBM Computational Science Division

2
Improving R performance

• Performance improvements
• Hardware (Number of cores etc.)
• Intel quad-core _at_2.4 Ghz Intel Q6600
• Compilers
• Intel versus GNU
• Compiler flags (unoptimized versus optimized)
• Libraries (BLAS)
• netlib BLAS, GotoBLAS2, Intel MKL, Intel MKL-SMP

3
Benchmark for R

• R-benchmark-25.R
• http//r.research.att.com/benchmarks/R-benchmark-2
  5.R
• Measures timings for
• B A A,
• C A/B
• Eigenvalues, Determinant, Cholesky, Inverse
  (BLAS)
• Needs SuppDists package
• ./Rscript --vanilla R-benchmark-25.R

4
Base R

• ./configure prefix/home/vsachde/R-install
• Source directory .
• Installation directory /home/vsachde/R-project
  /all-R/GNU-R/R-native-unoptimized
• C compiler gcc -stdgnu99 -g
  -O2
• Fortran 77 compiler gfortran -g -O
• C compiler g -g -O2
• Fortran 90/95 compiler gfortran -g -O
• Obj-C compiler
• Interfaces supported X11, tcltk
• External libraries readline
• Additional capabilities PNG, JPEG, TIFF,
  NLS, cairo
• Options enabled static R library,
  shared BLAS, R profiling, Java

Compiler flags
GNU Compilers
External libraries being used
5
Somewhat Optimized R

• export optim_flags-O3 -funroll-loops
  -ffast-math -marchcore2
• CC"gcc" CFLAGSoptim_flags CXX"g"
  CXXFLAGSoptim_flags F77"gfortran"
  FFLAGSoptim_flags FC"gfortran"
  FCFLAGSoptim_flags ./configure
  prefixinstalldir

C compiler gcc -stdgnu99 -O3
-funroll-loops -ffast-math -marchcore2 Fortran
77 compiler gfortran -O3 -funroll-loops
-ffast-math -marchcore2 C compiler
g -O3 -funroll-loops -ffast-math
-marchcore2 Fortran 90/95 compiler
gfortran -O3 -funroll-loops -ffast-math
-marchcore2

• Compilers can be changed by variables CC, CXX,
  F77
• CCicc CXXicpc F77ifort will use Intel
  compilers.

6
Linking external BLAS with R

• R uses unoptimized routines to do linear algebra
  if not linked with external BLAS.
• ./configure -with-blasltlocation of BLAS libgt
• Various sources of BLAS
• Netlib BLAS - Generic and unoptimized
• GotoBLAS2 Optimized and multi-threaded
• Intel MKL Optimized library from Intel
  (sequential)
• Intel MKL-SMP (Multi-threaded)
• Many others including ACML, Atlas.
• Performance of kernels change on different
  libraries used.

Tries to link the BLAS library
7
Linking external BLAS with R

• If everything goes well
• Source directory .
• Installation directory /home/vsachde/R-proje
  ct/all-R/GNU-R/R-netlib-blas
• C compiler gcc -stdgnu99 -O3
  -funroll-loops -ffast-math -marchcore2
• Fortran 77 compiler gfortran -O3
  -funroll-loops -ffast-math -marchcore2
• C compiler g -O3
  -funroll-loops -ffast-math -marchcore2
• Fortran 90/95 compiler gfortran -O3
  -funroll-loops -ffast-math -marchcore2
• Obj-C compiler
• Interfaces supported X11, tcltk
• External libraries readline,
  BLAS(generic)
• Additional capabilities PNG, JPEG, TIFF,
  NLS, cairo
• Options enabled static R library, R
  profiling, Java
• Recommended packages yes

BLAS was linked in properly
8
Linking external BLAS with R

• What does -with-blas do ?
• Link and run R with dgemm.
• configure28567 checking for dgemm_ in
  /home/vsachde/R-project/all-blas/GNU-blas/netlib-b
  las/libblas_GNU.a
• configure28588 gcc -stdgnu99 -o conftest -g
  -O2 -I/usr/local/include -L/usr/local/lib64
  conftest.c /home/vsachde/R-project/all-blas/GNU-bl
  as/netlib-blas/libblas_GNU.a -lgfortran -lm -ldl
  -lm gt5
• configure28595 result yes
• If the above linking step fails
• Installation wont fail, but BLAS will not be
  linked in.
• Summary at end wont show external BLAS linking.
• Search for dgemm in config.log and look for
  errors.
• Advice Compile static libraries as they are
  easier to link

9
Linking with different BLAS

• Netlib-BLAS
• Download source from netlib.org, unoptimized.
• GotoBLAS2
• Download from TACC website
• Optimized and multi-threaded
• Turn off CPU throttling to compile.
• Intel MKL
• Sequential and SMP
• Linking step is same for most BLASes except Intel
  libs

10
Linking with Intel MKL libs

• export MKLPATH/opt/intel/Compiler/11.1/072/mkl/li
  b/em64t/
• Intel MKL sequential
• --with-blas"-Wl,--start-group
  MKLPATH/libmkl_intel_lp64.a MKLPATH/libmkl_seque
  ntial.a MKLPATH/libmkl_core.a -Wl,--end-group
  -lpthread
• Intel MKL SMP
• --with-blas"-Wl,--start-group
  MKLPATH/libmkl_intel_lp64.a MKLPATH/libmkl_intel
  _thread.a MKLPATH/libmkl_core.a -Wl,--end-group
  -liomp5 -lpthread"

Intel MKL SMP and GotoBLAS2 should show
performance improvements in quad-core (run 4
threads)
11
Performance Single-thread BLAS
12
Performance BLAS
Performance went down by 15-20X through
compilers, compiler options and hardware (4
threads)
Revolution R uses Intel MKL-SMP
13
Results

• Generic R can be optimized for performance.
• Intel MKL libraries give best performance results
  with freely available GotoBLAS2 a close second.
• Experiment with LAPACK as well.
• Question How much is performance important for R
  users ?