Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Revise] add latency and bandwidth in benchmark #9

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

[Revise] add latency and bandwidth in benchmark #9

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 20 additions & 5 deletions FastMemcpy.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -77,9 +77,16 @@ void benchmark(int dstalign, int srcalign, size_t size, int times)
free(DATA1);
free(DATA2);

printf("result(dst %s, src %s): memcpy_fast=%dms memcpy=%d ms\n",
dstalign? "aligned" : "unalign",
srcalign? "aligned" : "unalign", (int)t2, (int)t1);
float lat1 = (float)t1 * 1000000 / times;
float lat2 = (float)t2 * 1000000 / times;

float bw1 = (float)size * times / 1024 / 1024 / 1024 / t1 * 1000;
float bw2 = (float)size * times / 1024 / 1024 / 1024 / t2 * 1000;

printf("result(dst %s, src %s): \tlatency(memcpy_fast:%.1fns memcpy=%.1fns) \tbandwith(memcpy_fast:%.1fGB/s memcpy=%.1fGB/s)\n",
dstalign? "aligned" : "unalign",
srcalign? "aligned" : "unalign",
lat2, lat1, bw2, bw1);
}


Expand Down Expand Up @@ -126,8 +133,16 @@ void random_bench(int maxsize, int times)
memcpy_fast(A + offset1, B + offset2, size);
}
t2 = gettime() - t2;
printf("benchmark random access:\n");
printf("memcpy_fast=%dms memcpy=%dms\n\n", (int)t2, (int)t1);

float lat1 = (float)t1 * 1000000 / times;
float lat2 = (float)t2 * 1000000 / times;

float bw1 = (float)maxsize / 2 * times / 1024 / 1024 / 1024 / t1 * 1000;
float bw2 = (float)maxsize / 2 * times / 1024 / 1024 / 1024 / t2 * 1000;

printf("benchmark random access(size=1~%d bytes, times=%d):\n", maxsize, times);
printf("result: \tlatency(memcpy_fast:%.1fns memcpy=%.1fns) \tbandwith(memcpy_fast:%.1fGB/s memcpy=%.1fGB/s)\n",
lat2, lat1, bw2, bw1);
}


Expand Down
24 changes: 19 additions & 5 deletions FastMemcpy_Avx.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -79,10 +79,16 @@ void benchmark(int dstalign, int srcalign, size_t size, int times)

free(DATA1);
free(DATA2);
float lat1 = (float)t1 * 1000000 / times;
float lat2 = (float)t2 * 1000000 / times;

printf("result(dst %s, src %s): memcpy_fast=%dms memcpy=%d ms\n",
dstalign? "aligned" : "unalign",
srcalign? "aligned" : "unalign", (int)t2, (int)t1);
float bw1 = (float)size * times / 1024 / 1024 / 1024 / t1 * 1000;
float bw2 = (float)size * times / 1024 / 1024 / 1024 / t2 * 1000;

printf("result(dst %s, src %s): \tlatency(memcpy_fast:%.1fns memcpy=%.1fns) \tbandwith(memcpy_fast:%.1fGB/s memcpy=%.1fGB/s)\n",
dstalign? "aligned" : "unalign",
srcalign? "aligned" : "unalign",
lat2, lat1, bw2, bw1);
}


Expand Down Expand Up @@ -129,8 +135,16 @@ void random_bench(int maxsize, int times)
memcpy_fast(A + offset1, B + offset2, size);
}
t2 = gettime() - t2;
printf("benchmark random access:\n");
printf("memcpy_fast=%dms memcpy=%dms\n\n", (int)t2, (int)t1);

float lat1 = (float)t1 * 1000000 / times;
float lat2 = (float)t2 * 1000000 / times;

float bw1 = (float)maxsize / 2 * times / 1024 / 1024 / 1024 / t1 * 1000;
float bw2 = (float)maxsize / 2 * times / 1024 / 1024 / 1024 / t2 * 1000;

printf("benchmark random access(size=1~%d bytes, times=%d):\n", maxsize, times);
printf("result: \tlatency(memcpy_fast:%.1fns memcpy=%.1fns) \tbandwith(memcpy_fast:%.1fGB/s memcpy=%.1fGB/s)\n",
lat2, lat1, bw2, bw1);
}


Expand Down