User:Manoj

Intel (2 x E5-2643 @ 3.30GHz)

Native FFT Library

Following library and flags were used:

MKLDIR    = $(HPC_MKL_DIR)
MKLLIBS   = -lmkl_intel_lp64 -lmkl_sequential -lmkl_core
MKLLIBDIR = $(HPC_MKL_DIR)/lib/intel64
FFTLIB = -lfftw3xf
INCS = -I$(MKLDIR)/include/fftw
FFT_OBJS = fftmpi.o fftmpi_map.o fftw3d.o fft3dlib.o
FFLAGS =  -free -names lowercase -assume byterecl
OFLAG  = -O2 -xsse2 -unroll-aggressive -warn general

As a first check, SIMD were changed and following is the result for MgMOS (For input file, please ask Charles Taylor or Manoj Srivastava):

SIMD Instruction	Time(s)
sse2	158
sse4.1	156
sse4.2	155
avx	155
ssse3	156

MKL FFTs (via FFTW wrappers)

Upon profiling the code, we found that the code spends most of its time in the FFT libraries, so the next step is to change FFTW libraries. Following changes were made:

FFT_OBJS = fftmpi_map.o fftmpiw.o fftw3d.o fft3dlib.o

(The change here is replacement of "fftmpi.o" of the original VASP makefile with "fftmpiw.o")

MKLDIR    = $(HPC_MKL_DIR)
MKLLIBS   = -lmkl_intel_lp64 -lmkl_sequential -lmkl_core
MKLLIBDIR = $(HPC_MKL_DIR)/lib/intel64
FFTLIB = -lfftw3xf
INCS = -I$(MKLDIR)/include/fftw
FFLAGS = -free -names lowercase -assume byterecl
OFLAG  = -O2 -xsse2 -unroll-aggressive -warn general

Upon making above changes, about 60% improvement on the run time of the code was found on the Intel machine (E5-2643 @ 3.30GHz). Following table depicts the run time variation with SIMD instruction sets:

SIMD Instruction	Time(s)
sse2	97
sse4.1	95
sse4.2	94
avx	94
ssse3	94

FFTW FFTs

We further compiled VASP by using FFT library from FFTW with following flags:

MKLDIR    = $(HPC_MKL_DIR)
MKLLIBS   = -lmkl_intel_lp64 -lmkl_sequential -lmkl_core
MKLLIBDIR = $(HPC_MKL_DIR)/lib/intel64
FFTWDIR = /apps/fftw/3.3.2
FFTLIB  = -L$(FFTWDIR)/lib -lfftw3
INCS = -I$(FFTWDIR)/include
FFT_OBJS = fftmpi_map.o fftmpiw.o fftw3d.o fft3dlib.o
FFLAGS =  -free -names lowercase -assume byterecl
OFLAG  = -O2 -xsse2 -unroll-aggressive -warn general

From our previous experience, we conclude that the performance of VASP does not depend substantially on the SIMD instruction sets, so for FFTW library, we only tried one set. Following is the result:

SIMD Instruction	Time(s)
sse2	118

AMD (2 x 6220 @ 3.0 GHz)

This machine has 16 cores, in numactl terminology 4 NUMA nodes with 4 cores on each nodes. As the result of VASP depends heavily on the choice of FFT libraries, we checked various performance of this machine with different FFTs, namely, FFT provided by VASP package, MKL, and FFTW. We build FFTW libraries with various flags to see if we could find a better choice for FFTs. The libraries and flags used to compile VASP are as follows (FFT libraries were changed depending on which FFT we want to use):

MKLDIR    = $(HPC_MKL_DIR)
MKLLIBS   = -lmkl_intel_lp64 -lmkl_sequential -lmkl_core
MKLLIBDIR = $(HPC_MKL_DIR)/lib/intel64
FFTWDIR = /apps/fftw/3.3.2
FFTLIB  = -L$(FFTWDIR)/lib -lfftw3
INCS = -I$(FFTWDIR)/include
FFT_OBJS = fftmpi_map.o fftmpiw.o fftw3d.o fft3dlib.o
FFLAGS =  -free -names lowercase -assume byterecl
OFLAG  = -O2 -xsse2 -unroll-aggressive -warn general

The results are summarized in the following table:

FFT Library	Shared L2-Cached time(s)	Exclusive L2-Cache time (s)	Notes
Naive	399
MKL	261	159
FFTW	333	217	1
FFTW	-	203	2
FFTW	-	210	3
FFTW	-	217	4
FFTW	-	213	5

Run Scheme	Native	MKL	FFTW	FFTW	FFTW	FFTW	FFTW
Shared L2-Cache time(s)	399	261	333	-	-	-	-
Exclusive L2-Cache time (s)	274	159	217	203	210	217	213
Notes	-	-	1	2	3	4	5

¹ Default compiler Flags were used to build FFT.
² CFLAG=-O3, FFLAG=-O3, -enable sse2
³ enable-mpi CFLAG=-O3, FFLAG=-O3, -enable sse2
⁴ CC='opencc -march=bdver1' F77='openf90 -march=bdver1' CFLAGS='-msse3 -msse4.1 -msse4.2 -msse4a -mfma4 -O2'  FFLAGS='-msse3 -msse4.1 -msse4.2 -msse4a -mfma4 -O2' --enable-fma --enable-mpi
⁵ FFLAGS/ CFLAGS="-OPT:Ofast -mavx -mfma4 -march=bdver1 -O3 -fomit-frame-pointer -LNO:simd=2 -WOPT:sib=on -LNO:prefetch=2:pf2=0 -CG:use_prefetchnta=on -LNO:prefetch_ahead=4-malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math"

Performance Comparison

Following is a summary of results for the test case of MgMOS ran on the machines with 8 processesors.

Server	Native	MKL	FFTW	FFTW
Intel	158	97	118	-
Intel (Scaled)	174	106	130	-
AMD (Shared L2 Cache)	399	261	-	-
AMD (Exc. L2 Cache)	274	159	-	203
AMD Shared/AMD Exc.	1.46	1.64	-	-
AMD Exc./Intel (scaled)	1.57	1.50	-	-
Notes	-	-	1	2

¹ Compiled by UFHPC (Charles Taylor or Craig Prescott)
² CFLAG=-O3, FFLAG=-O3, -enable sse2