-
Notifications
You must be signed in to change notification settings - Fork 2
/
slurm-matmul-without-tiling.sh
51 lines (38 loc) · 1.27 KB
/
slurm-matmul-without-tiling.sh
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
#!/bin/bash -l
# NOTE the -l flag!
#
# This is an example job file for a multi-core job.
# Note that all of the following statements below that begin
# with #SBATCH are actually commands to the SLURM scheduler.
# Please copy this file to your home directory and modify it
# to suit your needs.
#
# If you need any help, please email [email protected]
#
# Name of the job - You'll probably want to customize this.
#SBATCH -J matmul_test
# Standard out and Standard Error output files
#SBATCH -o matmul_test.output
#SBATCH -e matmul_test_error.output
#To send emails, set the adcdress below and remove one of the "#" signs.
##SBATCH --mail-user [email protected]
# notify on state change: BEGIN, END, FAIL or ALL
#SBATCH --mail-type=ALL
# Request 5 hours run time MAX, anything over will be KILLED
#SBATCH -t 5:0:0
# Put the job in the "work" partition and number of tasks as 1
# "work" is the default partition so it can be omitted without issue.
#SBATCH -p work -n 1
# Job memory requirements in MB
#SBATCH --mem=300
# Explicitly state you are a free user
#SBATCH --qos=free
#
# Your job script goes below this line.
#
# Load cuda module
module load cuda
# Compile cuda program
nvcc matrix_mul_without_tiling.cu -lcudart -o matmul
# Run the omp executable
srun matmul in_a.txt in_b.txt 4 5 4