ccf-ai-infra
/
GPUKernelContest
0
0
Fork 52
Code Issues Pull Requests 2 Packages Projects Releases Wiki Activity

赛题三个算子 (Reduce Sum, Sort Pair, Top-k Pair) 的优化实现 #13

Closed
wencong wants to merge 2 commits from wencong/GPUKernelContest:dev into main
pull from: wencong/GPUKernelContest:dev
merge into: ccf-ai-infra:main
Conversation 0 Commits 2 Files Changed 10 +2217 -8
wencong commented 2025-12-09 15:39:41 +08:00
First-time contributor
Copy Link

Benchmark 结果

实验配置

  • 实验机器：曦云 C500 - 16 GB * 1
  • 容器环境：GiteeAI pytorch2.4.0 maca 3.0.0.4 版本

实验结果

  • Reduce Sum 算子在小数据规模下(1000000)本算子吞吐量略低于参考实现；在大数据规模下(536870912 及以上)本算子吞吐量高于参考实现
  • Sort 算子在所有数据规模下吞吐量均显著高于参考实现
  • Top-k 算子在所有数据规模下吞吐量均显著高于参考实现
## Benchmark 结果 实验配置 - 实验机器：曦云 C500 - 16 GB * 1 - 容器环境：GiteeAI pytorch2.4.0 maca 3.0.0.4 版本 实验结果 - Reduce Sum 算子在小数据规模下(1000000)本算子吞吐量略低于参考实现；在大数据规模下(536870912 及以上)本算子吞吐量**高于**参考实现 - Sort 算子在**所有数据规模下**吞吐量均**显著**高于参考实现 - Top-k 算子在**所有数据规模下**吞吐量均**显著**高于参考实现
wencong added 1 commit 2025-12-09 15:39:41 +08:00
wencong added 1 commit 2025-12-09 16:28:43 +08:00
wencong closed this pull request 2025-12-09 16:29:28 +08:00

Pull request closed

Please reopen this pull request to perform a merge.
Sign in to join this conversation.
Reviewers
No reviewers
Labels
Clear labels
No items
No Label
Milestone
Clear milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
No Assignees
1 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: ccf-ai-infra/GPUKernelContest#13
No description provided.
Delete Branch "wencong/GPUKernelContest:dev"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?