[feat] enable hierarchical mc2 ops on A2 by default (#5300)
### What this PR does / why we need it?
Previously, it was necessary to set the environment variables
HCCL_INTRA_PCIE_ENABLE=1 and HCCL_INTRA_ROCE_ENABLE=0. This PR enables
hierarchical MC2 operations on A2 by default.
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
- vLLM version: release/v0.13.0
- vLLM main:
ad32e3e19c
Signed-off-by: hwhaokun <haokun0405@163.com>
Co-authored-by: realliujiaxu <realliujiaxu@163.com>
This commit is contained in:
@@ -958,14 +958,6 @@ def calculate_dp_buffer_size() -> int:
|
||||
return max(dp_buffer_size, _MIN_DP_BUFFER_SIZE)
|
||||
|
||||
|
||||
# Currently, when in A2, setting the environment variables HCCL_INTRA_PCIE_ENABLE=1
|
||||
# and HCCL_INTRA_ROCE_ENABLE=0 can reduce cross-machine communication traffic and
|
||||
# significantly improve communication performance of MC2 ops dispatch/combine.
|
||||
def is_hierarchical_communication_enabled():
|
||||
return (os.getenv("HCCL_INTRA_ROCE_ENABLE", "") == "0"
|
||||
and os.getenv("HCCL_INTRA_PCIE_ENABLE", "") == "1")
|
||||
|
||||
|
||||
def has_layer_idx(model_instance: torch.nn.Module) -> bool:
|
||||
if model_instance is None:
|
||||
return False
|
||||
|
||||
Reference in New Issue
Block a user