-
Notifications
You must be signed in to change notification settings - Fork 24
/
README.ISPASS-2009
139 lines (91 loc) · 4.58 KB
/
README.ISPASS-2009
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
The directories AES, BFS, CP, LPS, LIB, MUM, NN, NQU, RAY, STO, and WP contain
benchmarks used in the ISPASS 2009 paper on GPGPU-Sim. Their original sources
are listed below.
The instructions below describe how to build and run the benchmarks assuming
you are using GPGPU-Sim v3.x. This series of instructions is essentially the
procedure we follow when using GPGPU-Sim v3.x in our research at UBC.
Assuming all dependencies required by the benchmarks are satisifed, you can
build/run them by doing the following:
1. Install the NVIDIA CUDA Toolkit and then install/build the NVIDIA CUDA SDK
benchmarks. Note the CUDA Toolkit needs to be installed before you can build
the CUDA SDK. Building the CUDA SDK is required since some of the ISPASS 2009
benchmarks use a library created when building the CUDA SDK.
2. Open Makefile.ispass-2009 in your favorite text editor and set the
variables at the top of the file to suit your environment.
3. Define the following environment variables:
CUDA_INSTALL_PATH
NVIDIA_COMPUTE_SDK_LOCATION
The first, CUDA_INSTALL_PATH, should point to the directory you installed
the NVIDIA CUDA Toolkit (e.g., /usr/local/cuda).
The second, NVIDIA_COMPUTE_SDK_LOCATION, should point to the directory you
installed the NVIDIA GPU Computing SDK (e.g., ~/NVIDIA_GPU_Computing_SDK)
You must also ensure your PATH includes $CUDA_INSTALL_PATH/bin.
4. run "make -f Makefile.ispass-2009" in this directory. You should NOT need
to copy or move any files for this to work.
5. Verify binaries were generated for the benchmarks in ../bin/release/
(relative to the directory this file located in)
6. Type "source setup_environment" in the v3.x directory. This modifies your
LD_LIBRARY_PATH to use GPGPU-Sim for CUDA/OpenCL applications. If you have
not already done so, build GPGPU-Sim now.
7. Place a link to the configuration files (gpgpusim.config and the
interconnect configuration file) in the simulation run directory. You can
do this using the script "setup_config.sh" you will find in this directory,
which creates symbolic links. For example type:
./setup_config.sh GTX480
Aside: If later you want to change the configuration, you need to first run
./setup_config.sh --cleanup
then run ./setup_config.sh again with the configuration you want.
8. Run one of the applications by typing the command line in the
README.GPGPU-Sim in the benchmark directory. For example,
cd AES
sh README.GPGPU-Sim
9. When debugging the simulator, you should build GPGPU-Sim in debug mode:
cd $GPGPUSIM_ROOT
source setup_environment debug
make clean
make
Example of running the simulator in gdb (starting from this directory):
cd AES
gdb --args `cat README.GPGPU-Sim` # different steps required for WP
###############################################################################
# References #
###############################################################################
AES
S. A. Manavski. CUDA compatible GPU as an efficient hardware
accelerator for AES cryptography. In ICSPC 2007: Proc. of IEEE Int’l
Conf. on Signal Processing and Communication, pages 65–68, 2007.
BFS:
P. Harish and P. J. Narayanan. Accelerating Large Graph Algorithms
on the GPU Using CUDA. In HiPC, pages 197–208, 2007.
CP:
http://www.crhc.uiuc.edu/IMPACT/parboil.php.
DG:
T. C. Warburton. Mini Discontinuous Galerkin Solvers.
http://www.caam.rice.edu/˜timwar/RMMC/MIDG.html.
LPS:
M. Giles. Jacobi iteration for a Laplace discretisation on a 3D structured
grid. http://people.maths.ox.ac.uk/˜gilesm/hpc/NVIDIA/laplace3d.pdf
LIB:
M. Giles and S. Xiaoke. Notes on using the NVIDIA 8800 GTX graphics
card. http://people.maths.ox.ac.uk/˜gilesm/hpc/
MUM:
M. Schatz, C. Trapnell, A. Delcher, and A. Varshney. High-throughput
sequence alignment using Graphics Processing Units. BMC Bioinformatics,
8(1):474, 2007.
NN:
Billconan and Kavinguy. A Neural Network on GPU.
http://www.codeproject.com/KB/graphics/GPUNN.aspx.
NQU:
Pcchen. N-Queens Solver.
http://forums.nvidia.com/index.php?showtopic=76893
RAY:
Maxime. Ray tracing. http://www.nvidia.com/cuda.
STO:
S. Al-Kiswany, A. Gharaibeh, E. Santos-Neto, G. Yuan, and M. Ripeanu.
StoreGPU: exploiting graphics processing units to accelerate distributed
storage systems. In Proc. 17th Int’l Symp. on High Performance
Distributed Computing, pages 165–174, 2008.
WP:
J. Michalakes and M. Vachharajani. GPU acceleration of numerical
weather prediction. IPDPS 2008: IEEE Int’l Symp. on Parallel and
Distributed Processing, pages 1–7, April 2008.