Skip to content

[Perf] Use small max_num_batched_tokens for A100 #17885

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
May 11, 2025
+5 −1
Merged

[Perf] Use small max_num_batched_tokens for A100 #17885

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
e5530f0
use small max_num_batched_tokens for A100
KuntaiDu May 9, 2025
2ed6ffc
link PR number
KuntaiDu May 9, 2025
1c6c748
Merge remote-tracking branch 'upstream/main' into kuntai-fix-a100-perf
KuntaiDu May 9, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 5 additions & 1 deletion vllm/engine/arg_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1457,11 +1457,15 @@ def _set_default_args_v1(self, usage_context: UsageContext) -> None:
from vllm.platforms import current_platform
try:
device_memory = current_platform.get_device_total_memory()
device_name = current_platform.get_device_name().lower()
except Exception:
# This is only used to set default_max_num_batched_tokens
device_memory = 0

if device_memory >= 70 * GiB_bytes:
# NOTE(Kuntai): Setting large `max_num_batched_tokens` for A100 reduces
# throughput, see PR #17885 for more details.
# So here we do an extra device name check to prevent such regression.
if device_memory >= 70 * GiB_bytes and "a100" not in device_name:
# For GPUs like H100 and MI300x, use larger default values.
default_max_num_batched_tokens = {
UsageContext.LLM_CLASS: 16384,
Expand Down