CUDA (A, B) - Introduction (warming up!!)

A5 (10 marks)
Focus: CUDA (A, B) - Introduction (warming up!!)
Q1. [+3] Querying your GPU: In this question, you will run a simple query code to know the properties
and limits of your NVIDIA card. Locate the “CUDA Samples” folder on your hard disk (e.g.,
c:\ProgramData\NVIDIA Corporation\CUDA Samples). Navigate to “1_Utilities\deviceQuery”
subfolder. Then, Open the solution file (.sln) that matches your Visual Studio version. Once opened,
compile and run project (Ctrl + F5). Then, capture your answers and submit them as an image file
named A5_Q1.png.
While the above sample project provide detailed information, a simpler code (with less information)
is given below. Use the below code if you cannot find or run the CUDA sample project.
Note: When creating a new project, make sure to choose CUDA template. The file extension for your
CUDA program should be “cu”.
Marking guide: +3 for a screenshot with the required info
#include "cuda_runtime.h"
#include "device_launch_parameters.h"
#include <stdio.h>
int main(){
cudaDeviceProp prop;
int count;
cudaGetDeviceCount(&count);
for (int i = 0; i < count; i++) {
cudaGetDeviceProperties(&prop, i);
printf("----- General Information for device %d ---\n", i);
printf("Name: %s\n", prop.name);
printf("Compute capability: %d.%d\n", prop.major, prop.minor);
printf("Clock rate: %d\n", prop.clockRate);
printf("Device copy overlap: ");
printf(prop.deviceOverlap ? "Enabled\n" : "Disabled\n");
printf("Kernel execution timeout: ");
printf(prop.kernelExecTimeoutEnabled ? "Enabled\n" : "Disabled\n");
printf("----- Memory Information for device %d ---\n", i);
printf("Total global mem: %lu\n", prop.totalGlobalMem);
printf("Total constant Mem: %ld\n", prop.totalConstMem);
printf("Max mem pitch: %ld\n", prop.memPitch);
printf("Texture Alignment: %ld\n", prop.textureAlignment);
printf("----- MP Information for device %d ---\n", i);
printf("Multiprocessor count: %d\n", prop.multiProcessorCount);
printf("Shared mem per mp: %ld\n", prop.sharedMemPerBlock);
printf("Registers per mp: %d\n", prop.regsPerBlock);
printf("Threads in warp: %d\n", prop.warpSize);
printf("Max threads per block: %d\n", prop.maxThreadsPerBlock);
printf("Max thread dimensions: (%d, %d, %d)\n",
prop.maxThreadsDim[0], prop.maxThreadsDim[1],
prop.maxThreadsDim[2]);
printf("Max grid dimensions: (%d, %d, %d)\n",
prop.maxGridSize[0], prop.maxGridSize[1],
prop.maxGridSize[2]);
printf("\n");
}
return 0;
}
Q2. [+7] Simple CUDA code: consider this loop for initializing an array a:
conts int n = 10000000; //10 millions
for (i = 0; i < n; i++)
a[i] = (double)i / n;
Submit:
a) The serial implementation running on the CPU.
b) The CUDA implementation (1 thread per array element).
In both cases, add code to print the first and last 5 elements of the array to verify your code. (not
that you need to use the placeholder %.7f to print 7 digits after the decimal point.
Sample output:
a[0]: 0.0000000
a[1]: 0.0000001
a[2]: 0.0000002
a[3]: 0.0000003
a[4]: 0.0000004
...
a[9999995]: 0.9999995
a[9999996]: 0.9999996
a[9999997]: 0.9999997
a[9999998]: 0.9999998
a[9999999]: 0.9999999
Marking guide:
+2 for measuring the time of the parallel and serial code
+2 for the kernel function
+3 for launch configuration and properly calling the kernel
Submission Instructions
For this assignment, you need to do the following:
1- Compress the PNG file from Q1 and the source code file (i.e. the .cu file, not the whole project)
from Q2 into one zip folder and give a name to the zipped file that matches your ID (e.g.,
1234567.zip).
2- Submit the zipped file to Canvas.
Note that you can resubmit an assignment, but the new submission overwrites the old submission and
receives a new timestamp.