[fused_all_gather_scaled_matmul] support rowwise scaling #137805
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
[ghstack-poisoned]
This was referenced Oct 11, 2024
pytorch-bot
bot
added
the
oncall: distributed
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/137805
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 84a9c66 with merge base dae6007 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
yifuwang
changed the title
This PR add support for `A_scale` to be row-wise scale. The op can automatically detect whether the row-wise scale is sharded or replicated. When the row-wise scale is sharded, the op would all-gather the scale in a pipelined fashion. cc XilunWu H-Huang awgu kwen2501 wanchaol fegin fduwjj wz337 wconstab d4l3k c-p-i-o [ghstack-poisoned]
This was referenced Oct 12, 2024
pytorchmergebot
pushed a commit
that referenced
this pull request
Oct 15, 2024
```
Parallelization strategy: every rank issues independent compute
-> barrier -> p2p copy sequences on two streams. In addition to
computation/communication overlapping, the strategy allows for
computation/computation overlapping, greatly reducing
quantization inefficiency.
Ideally, stream activities would look like this ("b" for
barriers, "cp" for p2p copies):
[rank 0]
stream 0: [ chunk_producer ][b][ cp ][ chunk_producer ][b][ cp ]
stream 1: [ chunk_producer ][b][ cp ][ chunk_producer ][b][ cp ]
[rank 1]
stream 0: [ chunk_producer ][b][ cp ][ chunk_producer ][b][ cp ]
stream 1: [ chunk_producer ][b][ cp ][ chunk_producer ][b][ cp ]
Note that the barriers synchronize streams with the same ID
across ranks. They don't synchronize streams on the same rank.
Since the work on both streams is independent, there's no
guarantee that the chunk_producer from stream 0 or stream 1 will
be scheduled first. If there is a scheduling mismatch across
ranks, the barrier forces all ranks to wait for the slowest.
When scheduling mismatches occur among ranks, the stream
activities might look like this (note that p2p copies from
different streams cannot overlap with each other):
[rank 0]
stream 0: [ chunk_producer ][b ][ cp ][ chunk_producer ][b ][ cp ]
stream 1: [ chunk_producer ][b] [ cp ][ chunk_producer ][b] [ cp ]
[rank 1]
stream 0: [ chunk_producer ][b] [ cp ][ chunk_producer ][b] [ cp ]
stream 1: [ chunk_producer ][b ][ cp ][ chunk_producer ][b ][ cp ]
To prevent this, we need to ensure that the chunk_producer on
stream 1 gets scheduled first on every rank. Without access to
the underlying kernels, CUDA offers no API to control the
scheduling order of two independent, overlapping kernels. Our
solution is to issue a small sleep kernel in stream 0. The sleep
duration is insignificant, but having an extra task in stream 0
will almost guarantee that the chunk_producer on stream 1 gets
scheduled first. Once the first chunk_producer is scheduled in
the correct order, there's very little room for the scheduling
order of subsequent kernels to be inconsistent across ranks.
```
Currently, we perform stream synchronization to ensure scheduling order. The stream synchronization has no bearing on correctness, but prevents inconsistent scheduling orders across ranks.
Without the stream synchronization, ranks may have inconsistent scheduling order, and the barriers cause all ranks to wait for the slowest rank:
<img width="379" alt="image" src="https://github.com/user-attachments/assets/ffb97e76-7e19-4449-b121-83c32ec3e91d">
With stream synchronization, the inconsistent scheduling order issue is addressed, but we lose compute/compute overlapping (this is the state before this PR):
<img width="378" alt="image" src="https://github.com/user-attachments/assets/4cb76246-625f-4fc1-b49a-823ae46d3f23">
With this PR, we get both consistent scheduling order across ranks and compute/compute overlap:
<img width="327" alt="image" src="https://github.com/user-attachments/assets/51ab1bdc-4f60-46e0-b53c-6d208e2d4888">
Pull Request resolved: #137836
Approved by: https://github.com/weifengpy
ghstack dependencies: #137643, #137738, #137805
pytorchmergebot
pushed a commit
that referenced
this pull request
Oct 15, 2024
…37850) ``` Parallelization strategy: after each rank copies its shard into its local p2p buffer, every rank issues independent p2p copy -> shard_consumer sequences to two streams. In addition to computation/communication overlapping, the strategy allows for computation/computation overlapping, greatly reducing quantization inefficiency. Notation: - "mv" for the copy to local buffer - "cp" for p2p copies - "b" for barriers Constraints: - The GPU scheduler may or may not overlap "mv" with the first shard_consumer. - "cp" from different streams cannot overlap. Ideal scenario 0 - "mv" overlaps with the first shard_consumer: stream 0: [ shard_consumer ][ cp ][ shard_consumer ] stream 1: [ mv ][b][ cp ][ shard_consumer ] Ideal scenario 1 - "mv" is scheduled before the first shard_consumer: stream 0: [ shard_consumer ][ cp ][ shard_consumer ] stream 1: [ mv ][b][ cp ][ shard_consumer ] Suboptimal scenario 0 - "mv" is scheduled after the first shard_consumer: stream 0: [ shard_consumer ] [ cp ][ shard_consumer ] stream 1: [ mv ][b][ cp ][ shard_consumer ] Suboptimal scenario 0 - "b" is scheduled after the first shard_consumer: stream 0: [ shard_consumer ] [ cp ][ shard_consumer ] stream 1: [ mv ] [b][ cp ][ shard_consumer ] We haven't yet figured out a way to ensure "mv" and "b" are either overlapped with or scheduled before the first shard_consumer. Thus, to prevent suboptimal scenarios, we are giving up the chance to overlap "mv" and "b" with the first shard_consumer for now. ``` This PR improves the scheduling for mm kernels with high SM utilization. The GPU scheduler tends to not overlap local DtoD copies with such kernels, which leads to suboptimal scheduling. The following is an example of pipelining PyTorch's cutlass-based, row-wise scaling fp8 kernel: Before this PR: <img width="298" alt="image" src="https://github.com/user-attachments/assets/81e0a7f4-18ee-47c6-b258-04fdaca7a6a2"> With this PR: <img width="253" alt="image" src="https://github.com/user-attachments/assets/982de5a8-da1e-4a8f-b67e-c9c869b0a77f"> Pull Request resolved: #137850 Approved by: https://github.com/weifengpy ghstack dependencies: #137643, #137738, #137805, #137836
This was referenced Oct 15, 2024
jackzhxng
pushed a commit
that referenced
this pull request
Oct 16, 2024
This PR add support for `A_scale` to be row-wise scale. The op can automatically detect whether the row-wise scale is sharded or replicated. When the row-wise scale is sharded, the op would all-gather the scale in a pipelined fashion. Pull Request resolved: #137805 Approved by: https://github.com/weifengpy ghstack dependencies: #137643, #137738
jackzhxng
pushed a commit
that referenced
this pull request
Oct 16, 2024
```
Parallelization strategy: every rank issues independent compute
-> barrier -> p2p copy sequences on two streams. In addition to
computation/communication overlapping, the strategy allows for
computation/computation overlapping, greatly reducing
quantization inefficiency.
Ideally, stream activities would look like this ("b" for
barriers, "cp" for p2p copies):
[rank 0]
stream 0: [ chunk_producer ][b][ cp ][ chunk_producer ][b][ cp ]
stream 1: [ chunk_producer ][b][ cp ][ chunk_producer ][b][ cp ]
[rank 1]
stream 0: [ chunk_producer ][b][ cp ][ chunk_producer ][b][ cp ]
stream 1: [ chunk_producer ][b][ cp ][ chunk_producer ][b][ cp ]
Note that the barriers synchronize streams with the same ID
across ranks. They don't synchronize streams on the same rank.
Since the work on both streams is independent, there's no
guarantee that the chunk_producer from stream 0 or stream 1 will
be scheduled first. If there is a scheduling mismatch across
ranks, the barrier forces all ranks to wait for the slowest.
When scheduling mismatches occur among ranks, the stream
activities might look like this (note that p2p copies from
different streams cannot overlap with each other):
[rank 0]
stream 0: [ chunk_producer ][b ][ cp ][ chunk_producer ][b ][ cp ]
stream 1: [ chunk_producer ][b] [ cp ][ chunk_producer ][b] [ cp ]
[rank 1]
stream 0: [ chunk_producer ][b] [ cp ][ chunk_producer ][b] [ cp ]
stream 1: [ chunk_producer ][b ][ cp ][ chunk_producer ][b ][ cp ]
To prevent this, we need to ensure that the chunk_producer on
stream 1 gets scheduled first on every rank. Without access to
the underlying kernels, CUDA offers no API to control the
scheduling order of two independent, overlapping kernels. Our
solution is to issue a small sleep kernel in stream 0. The sleep
duration is insignificant, but having an extra task in stream 0
will almost guarantee that the chunk_producer on stream 1 gets
scheduled first. Once the first chunk_producer is scheduled in
the correct order, there's very little room for the scheduling
order of subsequent kernels to be inconsistent across ranks.
```
Currently, we perform stream synchronization to ensure scheduling order. The stream synchronization has no bearing on correctness, but prevents inconsistent scheduling orders across ranks.
Without the stream synchronization, ranks may have inconsistent scheduling order, and the barriers cause all ranks to wait for the slowest rank:
<img width="379" alt="image" src="https://github.com/user-attachments/assets/ffb97e76-7e19-4449-b121-83c32ec3e91d">
With stream synchronization, the inconsistent scheduling order issue is addressed, but we lose compute/compute overlapping (this is the state before this PR):
<img width="378" alt="image" src="https://github.com/user-attachments/assets/4cb76246-625f-4fc1-b49a-823ae46d3f23">
With this PR, we get both consistent scheduling order across ranks and compute/compute overlap:
<img width="327" alt="image" src="https://github.com/user-attachments/assets/51ab1bdc-4f60-46e0-b53c-6d208e2d4888">
Pull Request resolved: #137836
Approved by: https://github.com/weifengpy
ghstack dependencies: #137643, #137738, #137805
jackzhxng
pushed a commit
that referenced
this pull request
Oct 16, 2024
…37850) ``` Parallelization strategy: after each rank copies its shard into its local p2p buffer, every rank issues independent p2p copy -> shard_consumer sequences to two streams. In addition to computation/communication overlapping, the strategy allows for computation/computation overlapping, greatly reducing quantization inefficiency. Notation: - "mv" for the copy to local buffer - "cp" for p2p copies - "b" for barriers Constraints: - The GPU scheduler may or may not overlap "mv" with the first shard_consumer. - "cp" from different streams cannot overlap. Ideal scenario 0 - "mv" overlaps with the first shard_consumer: stream 0: [ shard_consumer ][ cp ][ shard_consumer ] stream 1: [ mv ][b][ cp ][ shard_consumer ] Ideal scenario 1 - "mv" is scheduled before the first shard_consumer: stream 0: [ shard_consumer ][ cp ][ shard_consumer ] stream 1: [ mv ][b][ cp ][ shard_consumer ] Suboptimal scenario 0 - "mv" is scheduled after the first shard_consumer: stream 0: [ shard_consumer ] [ cp ][ shard_consumer ] stream 1: [ mv ][b][ cp ][ shard_consumer ] Suboptimal scenario 0 - "b" is scheduled after the first shard_consumer: stream 0: [ shard_consumer ] [ cp ][ shard_consumer ] stream 1: [ mv ] [b][ cp ][ shard_consumer ] We haven't yet figured out a way to ensure "mv" and "b" are either overlapped with or scheduled before the first shard_consumer. Thus, to prevent suboptimal scenarios, we are giving up the chance to overlap "mv" and "b" with the first shard_consumer for now. ``` This PR improves the scheduling for mm kernels with high SM utilization. The GPU scheduler tends to not overlap local DtoD copies with such kernels, which leads to suboptimal scheduling. The following is an example of pipelining PyTorch's cutlass-based, row-wise scaling fp8 kernel: Before this PR: <img width="298" alt="image" src="https://github.com/user-attachments/assets/81e0a7f4-18ee-47c6-b258-04fdaca7a6a2"> With this PR: <img width="253" alt="image" src="https://github.com/user-attachments/assets/982de5a8-da1e-4a8f-b67e-c9c869b0a77f"> Pull Request resolved: #137850 Approved by: https://github.com/weifengpy ghstack dependencies: #137643, #137738, #137805, #137836
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack (oldest at bottom):
This PR add support for
A_scaleto be row-wise scale. The op can automatically detect whether the row-wise scale is sharded or replicated. When the row-wise scale is sharded, the op would all-gather the scale in a pipelined fashion.cc @XilunWu @H-Huang @awgu @kwen2501 @wanchaol @fegin @fduwjj @wz337 @wconstab @d4l3k @c-p-i-o