Pinned
342 contributions in the last year
Less
More
Activity overview
Contributed to
intel-isl/Open3D,
stotko/stdgpu,
intel-isl/open3d-cmake-find-package
and 3 other
repositories
Contribution activity
July 2021
Created 26 commits in 2 repositories
Created a pull request in intel-isl/Open3D that received 4 comments
Add DEPENDS option to import and build 3rdparty functions
Some readability cleanups:
Prefix internal functions for 3rdparty dependency handling with open3d_ for consistency.
Add DEPENDS argument to import…
+115
−103
•
4
comments
Opened 25 other pull requests in 2 repositories
stotko/stdgpu
14
merged
- unordered_map, unordered_set: Add support for custom allocators
- deque: Add support for custom allocators
- vector: Add support for custom allocators
- mutex: Add support for custom allocators
- bitset: Add support for custom allocators
- atomic: Add support for custom allocators
- memory: Add allocator-based create/destroy functions
- memory: Add register memory API
- General: Port HIP backend to CMake HIP language
- cmake: Explicitly enforce CUDA standard
- cmake: Update support for CMake 3.21
- src: Fix typo in allocator member names
- test: Upgrade googletest to 1.11.0
- scripts: Upgrade to thrust 1.13
intel-isl/Open3D
11
merged
- Fix inconsistencies in 3rdparty dependency handling
- Cleanup CUDA utilities
- Add missing include in Logging.h
- Cleanup CUDAScopedStream and add stream support to NPPImage
- Introduce CUDAScopedStream class for Multi-Stream Support
- Replace *_TARGET variable usage by Open3D::3rdparty_* targets
- Port remaining git submodules to URL downloading
- Simplify HTML copy command
- Port build submodules to URL downloading
- Introduce generic CachedMemoryManager
- Release CUDA cache in faiss index initialization
Reviewed 12 pull requests in 1 repository
intel-isl/Open3D 12 pull requests
- ParallelFor takes device as argument
- F64 and Robust Kernel Support in RegistrationICP
- Shortcuts for Dtypes
- Print on evey Malloc and Free for debugging
- Cleanup CUDAScopedStream and add stream support to NPPImage
- Ensure CUDAState is initialized before Cacher
- Introduce CUDAScopedStream class for Multi-Stream Support
- Introduce generic CachedMemoryManager
- Replace *_TARGET variable usage by Open3D::3rdparty_* targets
- ParallelFor and count # of physical CPU cores
- Port remaining git submodules to URL downloading
- Faster Sort for FixedRadiusSearch, HybridSearch
12
contributions
in private repositories
Jul 3 – Jul 22