dylanyunlong/project_6
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

[INFRA] Import NVIDIA/CCCL upstream as optimization reference library

Browse Source 
CCCL (CUDA C++ Core Libraries) provides:
- CUB: device/block/warp-level GPU primitives (reduce, scan, sort, topk)
- Thrust: high-level parallel algorithms (transform_reduce, sort, scan)
- libcudacxx: CUDA C++ standard library (atomics, barriers, memory)
- cudax: experimental features (memory resources, allocators)
- Tuning policies: per-SM hardware-specific algorithm parameters

Competition optimization vectors mapped to CCCL:
- Output TPS (83% weight): warp_reduce, block_reduce, device_topk
- Input TPS (14% weight): device_scan, block_load, prefetch
- Cache TPS (3% weight): prefix caching strategy patterns
- Memory (0.9 util): pooled/cached/buddy allocators

Source: https://github.com/NVIDIA/cccl (shallow clone, HEAD only)
License: Apache-2.0
This commit is contained in:
EngineX CI
2026-07-30 09:35:51 +00:00
parent b4d01f481e
commit 56fd68e7dd
8871 changed files with 1454674 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
cccl_upstream/thrust/internal/test/thrust.example.lambda.filecheck Normal file
View File

@@ -0,0 +1,10 @@
CHECK: SAXPY (functor method)
CHECK-NEXT: 2 * 1 + 1 = 3
CHECK-NEXT: 2 * 2 + 1 = 5
CHECK-NEXT: 2 * 3 + 1 = 7
CHECK-NEXT: 2 * 4 + 1 = 9
CHECK-NEXT: SAXPY (placeholder method)
CHECK-NEXT: 2 * 1 + 1 = 3
CHECK-NEXT: 2 * 2 + 1 = 5
CHECK-NEXT: 2 * 3 + 1 = 7
CHECK-NEXT: 2 * 4 + 1 = 9