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

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

cccl_upstream/cub/benchmarks/docker/.gitignore

@@ -0,0 +1 @@
+Dockerfile

cccl_upstream/cub/benchmarks/docker/recipe.py

@@ -0,0 +1,26 @@
+#!/usr/bin/env python
+
+import hpccm
+
+hpccm.config.set_container_format("docker")
+
+Stage0 = hpccm.primitives.baseimage(image="nvidia/cuda:12.2.0-devel-ubuntu22.04")
+Stage0 += hpccm.building_blocks.apt_get(
+    ospackages=[
+        "git",
+        "tmux",
+        "gcc",
+        "g++",
+        "vim",
+        "python3",
+        "python-is-python3",
+        "ninja-build",
+    ]
+)
+# Stage0 += hpccm.building_blocks.llvm(version='15', extra_tools=True, toolset=True)
+Stage0 += hpccm.building_blocks.cmake(eula=True, version="3.26.3")
+# Stage0 += hpccm.building_blocks.nsight_compute(eula=True, version='2023.1.1')
+Stage0 += hpccm.building_blocks.pip(
+    packages=["fpzip", "numpy", "pandas", "pynvml"], pip="pip3"
+)
+Stage0 += hpccm.primitives.environment(variables={"CUDA_MODULE_LOADING": "EAGER"})