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

Jump to bottom

[EPIC] Use programmatic dependent launch in all CUB algorithms #3115

Open
1 of 17 tasks
bernhardmgruber opened this issue Dec 10, 2024 · 1 comment
Open
1 of 17 tasks

[EPIC] Use programmatic dependent launch in all CUB algorithms #3115

bernhardmgruber opened this issue Dec 10, 2024 · 1 comment
Assignees
@bernhardmgruber
Labels
cub For all items related to CUB

Comments

@bernhardmgruber
Copy link
Contributor

bernhardmgruber commented Dec 10, 2024 •
edited
Loading

Programmatic dependent launch (PDL) allows subsequent kernels to skip a device-wide synchronization (executed on the host) and instead synchronize on the device at a source location specified by the programmer. This can eliminate bubbles inside a stream when a kernel is finishing until the next kernel can ramp up.

We should evaluate where we can use this feature in CUB and then add it wherever it shows a benefit.

  • DeviceAdjacentDifference
  • DeviceCopy
  • DeviceFor
  • DeviceHistogram
  • DeviceMemcpy
  • PR: Use programmatic dependent launch in CUB merge sort #3114
  • DeviceMerge
  • DevicePartition
  • DeviceRadixSort
  • DeviceReduce
  • DeviceRunLengthEncode
  • DeviceScan
  • DeviceSegmentedRadixSort
  • DeviceSegmentedReduce
  • DeviceSegmentedSort
  • DeviceSelect
  • DeviceTransform
@bernhardmgruber bernhardmgruber self-assigned this Dec 10, 2024
@github-project-automation github-project-automation bot moved this to Todo in CCCL Dec 10, 2024
@bernhardmgruber bernhardmgruber added the cub For all items related to CUB label Dec 10, 2024
@bernhardmgruber bernhardmgruber changed the title Use programmatic dependent launch in all CUB algorithms [EPIC] Use programmatic dependent launch in all CUB algorithms Dec 10, 2024
@bernhardmgruber bernhardmgruber mentioned this issue Dec 10, 2024
Merged
@gevtushenko
Copy link
Collaborator

Let's address #3131 before starting work on next algorithms from the list

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

@bernhardmgruber bernhardmgruber

Labels
cub For all items related to CUB
Projects
CCCL
Status: Todo
Milestone
No milestone
Development

No branches or pull requests
2 participants
@bernhardmgruber @gevtushenko