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

Need for parallel sorting algorithms #101

Open
poulson opened this issue Oct 27, 2015 · 1 comment
Open

Need for parallel sorting algorithms #101

poulson opened this issue Oct 27, 2015 · 1 comment
Labels
enhancement

Comments

@poulson
Copy link
Member

poulson commented Oct 27, 2015

Several routines in Elemental (e.g., sorting eigenpairs and singular triplets, and computing medians) requires sorting distributed data structures. It would be worthwhile for Elemental to incorporate (and/or implement) distributed selection sort and bitonic sort.
@poulson
Copy link
Member Author

poulson commented Jul 31, 2016

As mentioned in Issue #7, https://patflick.github.io/mxx/ might be a good solution for distributed sample sort.

benson31 added a commit to benson31/Elemental that referenced this issue Jul 31, 2020
@benson31
Support HIP/ROCm backends for GPUs (elemental#101)
d2feee8 
* Add hydrogen error handling mechanisms
* new cuda management infrastructure
* everything in rocm compiles i think. linker issues pending
* remove override decoration from Element/BlockMatrix functions
* patch for finding rocblas; not sure if this is strictly necessary any more
* forward kernel arguments by reference
* a few tweaks to the CMakeLists
* Make sure ROCm and CUDA aren't enabled at the same time.
* correct a discrepancy in hipMemcpy2DAsync semantics
* clean up HAVE_CUDA macro usage; streamline copy syntax
* use nonblocking stream; clean up the mempool
* straggler HAVE_CUDA use in include tree
* preprocessor macro cleanup in blaslike tests
* Remove debugging print statements
* add short-circuit returns to copy/fill routines when size is zero
* some cleanup
* fix some new rocm issues
* update aluminum version number
* update version number
* remove some unneeded CMake
* revert changes related to the hip override bug
* add support for hipCUB and generalize cublas tensor option
* fix annoying clang warnings (that GCC _should_ throw, too, but it doesn't)
* address some review comments
* fix use of streams that should have been SyncInfos
* Clean up device library functions
* cleanup timer nonsense in Gemm test
* fix some hipCUB linkage
* Apply suggestions from code review
Co-authored-by: Tim Moon <[email protected]>
* Apply suggestions from code review
Co-authored-by: Tim Moon <[email protected]>
* remove unneeded metafunction. DiHydrogen has a cleaner implementation anyway.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
enhancement
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@poulson