Skip to content
This repository has been archived by the owner on Feb 18, 2024. It is now read-only.

Improved performance of filter performance via Simd selection [3x] #871

Merged
merged 2 commits into from
Feb 28, 2022
Merged

Improved performance of filter performance via Simd selection [3x] #871

merged 2 commits into from
Feb 28, 2022

Conversation

sundy-li
Copy link
Collaborator

@sundy-li sundy-li commented Feb 28, 2022 •
edited
Loading

Benchmarks:

Before:

filter f32              time:   [269.62 us 269.88 us 270.12 us]                       
filter f32 high selectivity                                                                            
                        time:   [243.83 us 244.41 us 244.95 us]
filter f32 nonull       time:   [108.89 us 108.95 us 109.03 us]                              
filter f32 nonull high selectivity                                                                             
                        time:   [18.886 us 18.909 us 18.933 us]

Now:

filter f32            time:   [180.22 us 180.34 us 180.45 us]                       
filter f32 high selectivity                                                                            
                        time:   [94.458 us 94.495 us 94.540 us]
			
filter f32 nonull       time:   [58.747 us 58.775 us 58.809 us]                              
filter f32 nonull high selectivity                                                                             
                        time:   [10.456 us 10.477 us 10.494 us]

@codecov
Copy link

codecov bot commented Feb 28, 2022 •
edited
Loading

Codecov Report

Merging #871 (2b06aa2) into main (e545ae5) will decrease coverage by 0.00%.
The diff coverage is 64.70%.

Impacted file tree graph

@@            Coverage Diff             @@
##             main     #871      +/-   ##
==========================================
- Coverage   71.50%   71.49%   -0.01%     
==========================================
  Files         335      335              
  Lines       17950    17969      +19     
==========================================
+ Hits        12835    12847      +12     
- Misses       5115     5122       +7
Impacted Files Coverage Δ
benches/filter_kernels.rs 0.00% <ø> (ø)
src/compute/filter.rs 61.83% <47.82%> (-4.03%) ⬇️
src/types/bit_chunk.rs 100.00% <100.00%> (ø)
src/compute/arithmetics/time.rs 26.60% <0.00%> (+0.91%) ⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update e545ae5...2b06aa2. Read the comment docs.

@sundy-li sundy-li changed the title Simd selector for filter kenerl performance [10X] Simd selector for filter kenerl performance [3X] Feb 28, 2022
@sundy-li sundy-li marked this pull request as ready for review February 28, 2022 04:50
@sundy-li sundy-li mentioned this pull request Feb 28, 2022
Merged
@sundy-li sundy-li changed the title Simd selector for filter kenerl performance [3X] Simd selection for filter kenerl performance [3X] Feb 28, 2022
jorgecarleitao
jorgecarleitao reviewed Feb 28, 2022
View reviewed changes
Copy link
Owner

@jorgecarleitao jorgecarleitao left a comment

Choose a reason for hiding this comment

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

Amazing PR, @sundy-li ! Left a small number of comments but it is otherwise ready imo.

src/types/bit_chunk.rs Show resolved Hide resolved
Cargo.toml Show resolved Hide resolved
src/types/bit_chunk.rs Show resolved Hide resolved
src/types/bit_chunk.rs Show resolved Hide resolved
@jorgecarleitao jorgecarleitao changed the title Simd selection for filter kenerl performance [3X] Improved performance of filter performance via Simd selection [3x] Feb 28, 2022
@jorgecarleitao jorgecarleitao merged commit 5d0db54 into jorgecarleitao:main Feb 28, 2022
rapiz1
rapiz1 reviewed Mar 1, 2022
View reviewed changes
Cargo.toml
@@ -249,6 +249,7 @@ harness = false
name = "comparison_kernels"
harness = false


Copy link

Choose a reason for hiding this comment

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

A blank line slipped in

@jorgecarleitao jorgecarleitao mentioned this pull request Mar 1, 2022
Closed
sydduckworth pushed a commit to mindx/arrow2 that referenced this pull request Mar 2, 2022
@sundy-li
Improved performance of filter performance via Simd selection [3x] (j…
e1c8a04 
…orgecarleitao#871)
sydduckworth pushed a commit to mindx/arrow2 that referenced this pull request Mar 2, 2022
@sundy-li
Improved performance of filter performance via Simd selection [3x] (j…
4fbcfa2 
…orgecarleitao#871)
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Reviewers

@rapiz1 rapiz1 rapiz1 left review comments

@jorgecarleitao jorgecarleitao jorgecarleitao left review comments

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@sundy-li @jorgecarleitao @rapiz1