Skip to content

[CUDA][Performance] Add radix select implementation for efficient partition operations #3117

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

Open
Lyxot wants to merge 18 commits into
base: main
Choose a base branch
from
+1,213 −3
Open

[CUDA][Performance] Add radix select implementation for efficient partition operations #3117

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
18 commits
Select commit Hold shift + click to select a range
e2f1b88
Init CUDA radix select
Lyxot Feb 8, 2026
8084da7
update fallback strategy
Lyxot Feb 8, 2026
971b60d
format code
Lyxot Feb 8, 2026
e4c4306
cuda: fix radix partition row indexing and 64-bit strides
Lyxot Feb 8, 2026
1e523c1
cuda: remove radix-partition rank cap and use dynamic NC metadata
Lyxot Feb 8, 2026
c87e3c8
unify radix partition kernels to reduce code duplication
Lyxot Feb 9, 2026
ccf39c6
implement dynamic shared memory for radix partition small kernel
Lyxot Feb 9, 2026
cbefde1
introduce multi-block-per-row and multi-row-per-block
Lyxot Feb 9, 2026
e01761f
fix: make radix select tie-order deterministic
Lyxot Feb 9, 2026
10307f4
update benchmark scripts
Lyxot Feb 9, 2026
9eae0ca
fix: canonicalize signed zero in CUDA radix keys for deterministic ra…
Lyxot Feb 9, 2026
88184ce
fix: make radix select tie-order deterministic for small kernel
Lyxot Feb 10, 2026
62fe754
fall back to merge sort when radix nc_dim > MAX_NDIM
Lyxot Feb 10, 2026
041f207
tune radix-select small-kernel launch and radix width
Lyxot Feb 15, 2026
65dfbc3
remove multi-row-per-block
Lyxot Feb 15, 2026
fd20a11
replace the fixed argpartition small-kernel cutoff with a runtime check
Lyxot Feb 16, 2026
d9a85a0
remove large kernel
Lyxot Mar 2, 2026
5fc2113
remove runtime fit routing
Lyxot Mar 2, 2026
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
Loading