Magma 1.2.1 installation

[yingcai]

I am trying to install magma-1.2.1 on my Linux Ubuntu workstation.
[see my make.inc at the end]

My graphic card is

VGA compatible controller: nVidia Corporation GT200GL [NVIDIA Quadro FX 3800] (rev a1)

I am wondering in the "make.inc" file, should I use
GPU_TARGET = Fermi
or
GPU_TARGET = Tesla
?

I am using 'icc' and 'ifort' and INTEL MKL lapack & blas since I installed parallel_studio_xe_2013_intel64 (MKL is part of it and I set its path in $MKLROOT) on my computer.
what's the meaning of "-lguide" in the following line? I cannot find this library.

my "make.inc" file:
#######################################
GPU_TARGET = Fermi
GPU_TARGET = Tesla

CC = gcc
CC = icc
NVCC = nvcc
FORT = ifort

ARCH = ar
ARCHFLAGS = cr
RANLIB = ranlib

OPTS = -O3 -DADD_
F77OPTS = -O3 -DADD_ -cpp
FOPTS = -O3 -DADD_ -cpp
NVOPTS = --compiler-options -fno-strict-aliasing -DUNIX -O3 -DADD_
LDOPTS = -fPIC -nofor_main -Xlinker -zmuldefs

#LIB = -lmkl_intel_lp64 -lmkl_intel_thread -lmkl_lapack -lmkl_core -lguide -lpthread -lcublas -lcudart -lm
LIB = -L$(MKLROOT)/lib/intel64 -lmkl_intel_lp64 -lmkl_intel_thread -lmkl_lapack95_lp64 -lmkl_core $(INTEL)/lib/intel64/libiomp5.a -lpthread -lcublas -lcudart -lm

CUDADIR = /scratch1/local/cuda

LIBDIR = -L$(MKLROOT)/lib/intel64 \
-L$(CUDADIR)/lib64
INC = -I$(CUDADIR)/include

Yingcai

[mgates3]

According to the Nvidia site
http://developer.nvidia.com/cuda/cuda-gpus
the Quadro FX 3800 is compute capability 1.3, so use Tesla. (This is a misnomer; by Tesla we really mean pre-Fermi.)

In some (older?) versions of MKL there is the -lguide library. If it is not needed, just remove it. You can lookup the required MKL libraries in the MKL documentation on Intel's website.

-mark