https://github.com/xmas7/cudampi

A large hybrid CPU/GPU sorting network using CUDA and MPI. The sorting network uses a standard Quicksort for CPUs and a custom Bitonic Sort for GPUs. These two algorithms were the fastest in a number of prior benchmarks.
https://github.com/xmas7/cudampi

cpu cuda gpu hybrid mpi network

Last synced: about 2 months ago
JSON representation

A large hybrid CPU/GPU sorting network using CUDA and MPI. The sorting network uses a standard Quicksort for CPUs and a custom Bitonic Sort for GPUs. These two algorithms were the fastest in a number of prior benchmarks.

• Host: GitHub
• URL: https://github.com/xmas7/cudampi
• Owner: xmas7
• License: lgpl-3.0
• Created: 2022-09-02T03:29:58.000Z (over 2 years ago)
• Default Branch: master
• Last Pushed: 2022-09-02T03:30:31.000Z (over 2 years ago)
• Last Synced: 2025-02-01T09:28:03.711Z (4 months ago)
• Topics: cpu, cuda, gpu, hybrid, mpi, network
• Language: Shell
• Homepage:
• Size: 1.17 MB
• Stars: 2
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
CUDAMPI

=======

A large hybrid CPU/GPU sorting network using CUDA and MPI.

The sorting network uses a standard Quicksort for CPUs and a custom

Bitonic Sort for GPUs. These two algorithms were the fastest

in a number of prior benchmarks.

We execute the first step of a bucketsort algorithm to presort the data.

A bucket only contains numbers in a given range.

We put each number into its corresponding bucket. This can be done in parallel.

Now each bucket can be sorted on either a CPU or a GPU.

The sorting network uses the filesystem as a process management solution.

Therefore no explicit locks are required.

MPI/IO is used to write to the filesystem in parallel.

The result is a number of sorted files.