Skip to content

Conversation

@frankdjx
Copy link
Contributor

Add benchmark for aggregate min/max compute kernels

Signed-off-by: Frank Du frank.du@intel.com

@github-actions
Copy link

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't think it's important to benchmark the count kernel. It's a trivial O(1) kernel.

frankdjx and others added 2 commits August 12, 2020 19:11
Also add count.

Signed-off-by: Frank Du <frank.du@intel.com>
Copy link
Member

@pitrou pitrou left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you. Example output (AMD Ryzen, AVX2):

MinMaxKernelFloat/1048576/10000         222 us          222 us         3159 bytes_per_second=4.40764G/s null_percent=0.01 size=1048.58k
MinMaxKernelFloat/1048576/100           233 us          233 us         3002 bytes_per_second=4.19139G/s null_percent=1 size=1048.58k
MinMaxKernelFloat/1048576/10            399 us          399 us         1759 bytes_per_second=2.45046G/s null_percent=10 size=1048.58k
MinMaxKernelFloat/1048576/2             934 us          934 us          726 bytes_per_second=1071.02M/s null_percent=50 size=1048.58k
MinMaxKernelFloat/1048576/1             339 us          339 us         1973 bytes_per_second=2.87805G/s null_percent=100 size=1048.58k
MinMaxKernelFloat/1048576/0             187 us          187 us         3723 bytes_per_second=5.21389G/s null_percent=0 size=1048.58k

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants