mindspore-ai / hyper-parallel
A SuperPod-Affinity Training Acceleration Framework.
Why this job: works in the areas we cover, but does not say which job it does
What it does
HyperParallel provides distributed parallel acceleration for Ascend SuperPoDs. Designed for resource pooling, peer-to-peer architectures, hierarchical network topologies, and FP8 low-precision formats, it spans cluster-level distributed parallelism and on-chip multicore parallelism. It supports pooled management of CPU DRAM and NPU HBM, topology-aware scheduling, communication path planning, and FP8 mixed-precision training.
Get started
No single install command. Read the steps in the README (opens in new tab)
- Maintenance
- Active
- Last commit today
- Stars
- 212
- +95 this week
- Licence
- Apache-2.0
- Allows business use
- Downloads for
- None listed
- No release files
- Latest version
- v1.1.0
- Released yesterday
- Built with
- Python
- 97% of the code
Is it looked after?
- Contributors
- 24
- Last 90 days
- Issues
- 37% closed
- 503 still open
- Pull requests
- 302 merged
- 179 waiting
- Releases
- 1
- 1 in the last year
- Contributing guide (opens in new tab)
- No security policy
- No code of conduct
Similar repos
| Repo | Stars | Last commit | Licence | Downloads for |
|---|---|---|---|---|
| microsoft / nnscaler | 136 | 17 days ago | MIT | Linux |
| learning-at-home / hivemind | 2,552 | 10 months ago | MIT | No downloads |
| meta-pytorch / torchx | 429 | 45 days ago | Custom | No downloads |
More: languages, topics and project record
Languages
- Python 97%
- C++ 2%
- Shell <1%
- CMake <1%
Topics
- ai-infrastructure
- llm
- omni
- distributed-training
- pytorch
Project record
- JobOther AI tools
- StatusListed
- Commits2,113
- Forks110
- Watching4
- Stars in 28 days+134
- CreatedFeb 27, 2026
- First seen by usNov 7, 2026
- CheckedNov 12, 2026