This repository has been archived by the owner on Jan 10, 2025. It is now read-only.
Improves Address Translation Speed #131
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
... by using the eytzinger order instead of performing a binary search on the
regionsin ascending order.Additionally, the keys are copied into a dense vector without any gaps or stride, for better utilization of the cache.
Benchmarks
Overall, the performance benchmarks indicate a ~ 30% speed up in address translation.
Binary Search, Regions Directly
test bench_gapped_randomized_access_with_1024_entries ... bench: 18 ns/iter (+/- 7)
test bench_mapping_with_1024_entries ... bench: 18 ns/iter (+/- 3)
test bench_randomized_access_with_0001_entry ... bench: 20 ns/iter (+/- 4)
test bench_randomized_access_with_1024_entries ... bench: 74 ns/iter (+/- 14)
test bench_randomized_mapping_access_with_0004_entries ... bench: 28 ns/iter (+/- 5)
test bench_randomized_mapping_access_with_0016_entries ... bench: 41 ns/iter (+/- 9)
test bench_randomized_mapping_access_with_0064_entries ... bench: 52 ns/iter (+/- 10)
test bench_randomized_mapping_access_with_0256_entries ... bench: 62 ns/iter (+/- 8)
test bench_randomized_mapping_access_with_1024_entries ... bench: 70 ns/iter (+/- 12)
test bench_randomized_mapping_with_1024_entries ... bench: 18 ns/iter (+/- 9)
Binary Search, Dense Keys
test bench_gapped_randomized_access_with_1024_entries ... bench: 19 ns/iter (+/- 9)
test bench_mapping_with_1024_entries ... bench: 18 ns/iter (+/- 3)
test bench_randomized_access_with_0001_entry ... bench: 18 ns/iter (+/- 7)
test bench_randomized_access_with_1024_entries ... bench: 77 ns/iter (+/- 11)
test bench_randomized_mapping_access_with_0004_entries ... bench: 28 ns/iter (+/- 5)
test bench_randomized_mapping_access_with_0016_entries ... bench: 42 ns/iter (+/- 10)
test bench_randomized_mapping_access_with_0064_entries ... bench: 52 ns/iter (+/- 10)
test bench_randomized_mapping_access_with_0256_entries ... bench: 63 ns/iter (+/- 11)
test bench_randomized_mapping_access_with_1024_entries ... bench: 73 ns/iter (+/- 16)
test bench_randomized_mapping_with_1024_entries ... bench: 17 ns/iter (+/- 2)
Eytzinger Heap, Regions Directly
test bench_gapped_randomized_access_with_1024_entries ... bench: 23 ns/iter (+/- 4)
test bench_mapping_with_1024_entries ... bench: 40 ns/iter (+/- 7)
test bench_randomized_access_with_0001_entry ... bench: 22 ns/iter (+/- 4)
test bench_randomized_access_with_1024_entries ... bench: 50 ns/iter (+/- 10)
test bench_randomized_mapping_access_with_0004_entries ... bench: 29 ns/iter (+/- 5)
test bench_randomized_mapping_access_with_0016_entries ... bench: 31 ns/iter (+/- 5)
test bench_randomized_mapping_access_with_0064_entries ... bench: 37 ns/iter (+/- 7)
test bench_randomized_mapping_access_with_0256_entries ... bench: 43 ns/iter (+/- 6)
test bench_randomized_mapping_access_with_1024_entries ... bench: 45 ns/iter (+/- 21)
test bench_randomized_mapping_with_1024_entries ... bench: 40 ns/iter (+/- 6)
Eytzinger Heap, Dense Keys
test bench_gapped_randomized_access_with_1024_entries ... bench: 22 ns/iter (+/- 3)
test bench_mapping_with_1024_entries ... bench: 35 ns/iter (+/- 7)
test bench_randomized_access_with_0001_entry ... bench: 22 ns/iter (+/- 4)
test bench_randomized_access_with_1024_entries ... bench: 45 ns/iter (+/- 8)
test bench_randomized_mapping_access_with_0004_entries ... bench: 26 ns/iter (+/- 13)
test bench_randomized_mapping_access_with_0016_entries ... bench: 29 ns/iter (+/- 7)
test bench_randomized_mapping_access_with_0064_entries ... bench: 35 ns/iter (+/- 6)
test bench_randomized_mapping_access_with_0256_entries ... bench: 40 ns/iter (+/- 8)
test bench_randomized_mapping_access_with_1024_entries ... bench: 45 ns/iter (+/- 7)
test bench_randomized_mapping_with_1024_entries ... bench: 35 ns/iter (+/- 5)