Files
pytorch/torch
Ke Wen 5c827a4133 [SymmMem] Multi-root tile reduction (#164757)
Stack from [ghstack](https://github.com/ezyang/ghstack/tree/0.12.0) (oldest at bottom):

Perform multiple tile reductions concurrently, with each tile reduced to a separate root.

- The number of concurrent reductions can be smaller than world size, i.e. roots can be a subset of all ranks. But all ranks are still required to call into this API.

- Currently supports NVLink SHARP scope only.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164757
Approved by: https://github.com/weifengpy, https://github.com/fegin
ghstack dependencies: #162243
2025-10-08 17:28:00 +00:00
..
2025-10-08 14:23:38 +00:00
2025-10-08 14:23:38 +00:00
2025-10-04 03:40:32 +00:00
2025-10-08 14:23:38 +00:00
2025-10-08 02:30:57 +00:00
2025-10-08 14:23:38 +00:00
2025-10-08 02:30:57 +00:00
2025-10-08 02:30:57 +00:00
2025-10-08 07:27:17 +00:00
2025-10-08 02:30:57 +00:00
2025-09-29 14:49:19 +00:00
2025-10-08 07:27:17 +00:00
2025-10-08 02:30:57 +00:00
2025-10-08 02:30:57 +00:00
2025-10-08 07:27:17 +00:00
2025-10-08 02:30:57 +00:00
2025-04-27 09:56:42 +00:00
2025-10-08 14:23:38 +00:00
2025-10-08 02:30:57 +00:00