Skip to content

Navigation Menu

Appearance settings

View all features
- BY COMPANY SIZE
  Enterprises
  Small and medium teams
  Startups
  Nonprofits
- BY USE CASE
  App Modernization
  DevSecOps
  DevOps
  CI/CD
  View all use cases
- BY INDUSTRY
  Healthcare
  Financial services
  Manufacturing
  Government
  View all industries
View all solutions
- EXPLORE BY TOPIC
  AI
  Software Development
  DevOps
  Security
  View all topics
- EXPLORE BY TYPE
  Customer stories
  Events & webinars
  Ebooks & reports
  Business insights
  GitHub Skills
- SUPPORT & SERVICES
  Documentation
  Customer support
  Community forum
  Trust center
  Partners
View all resources
- COMMUNITY
  GitHub SponsorsFund open source developers
- PROGRAMS
  Security Lab
  Maintainer Community
  GitHub Stars
  Archive Program
- REPOSITORIES
  Topics
  Trending
  Collections
- ENTERPRISE SOLUTIONS
  Enterprise platformAI-powered developer platform
- AVAILABLE ADD-ONS
  GitHub Advanced SecurityEnterprise-grade security features
  Copilot for BusinessEnterprise-grade AI features
  Premium SupportEnterprise-grade 24/7 support
Pricing

Appearance settings

You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session.

Dismiss alert

Uh oh!

There was an error while loading. Please reload this page.

sgl-project / sglang Public

Notifications You must be signed in to change notification settings
Fork 8.9k
Star 36.1k

Code
Issues 903
Pull requests 4.5k
Discussions
Actions
Security and quality
Insights

Additional navigation options

Code
Issues
Pull requests
Discussions
Actions
Security and quality
Insights

[AMD] Perf Kimi-K3 low concurrency optimizations - #35287

#35287

Open

kkHuang-amd wants to merge 8 commits into

sgl-project:mainsgl-project/sglang:mainfrom

HaiShaw:perf/k3-gfx950-independent-fusions-cleanHaiShaw/sglang:perf/k3-gfx950-independent-fusions-cleanCopy head branch name to clipboard

Conversation Commits8 (8)Checks Files changed

Open

[AMD] Perf Kimi-K3 low concurrency optimizations#35287
kkHuang-amd wants to merge 8 commits into
sgl-project:mainfrom
HaiShaw:perf/k3-gfx950-independent-fusions-clean

Merge branch 'main' into perf/k3-gfx950-independent-fusions-clean

073250d

Select commit

Loading

Failed to load commit list.

Uh oh!

There was an error while loading. Please reload this page.

Sign in for the full log view

PR Test (sgl-router) on: pull_request 1

gate

tier-1 — lint

tier-2 — build + test

tier-3 — docker (placeholder)

tier-3 — k8s integration

tier-3 — e2e

finish

PR Test Extra on: pull_request 3

check-changes / check-changes

call-gate / pr-gate

sgl-kernel-build-wheels

rust-ext-build

extra-a-test-1-gpu-small

extra-a-test-1-gpu-large

extra-a-test-2-gpu-large

extra-b-test-4-gpu-h100

extra-b-test-4-gpu-b200

extra-b-test-8-gpu-h200

pr-test-extra-finish

notify-pr-states

Lint on: pull_request 2

lint

PR Test (Xeon) on: pull_request 3

check-changes

pr-gate / pr-gate

build-test (xeon-gnr, base-b-test-cpu)

PR Test Extra (AMD) on: pull_request 2

call-gate / pr-gate

extra-a-test-1-gpu-small-amd${{ ((inputs.rocm_version && (inputs.rocm_version != 'rocm724') && format('-{0}', inputs.rocm_version)) || '') }} (linux-${{ (inputs.runner_arch || 'mi300') }}-1gpu-sglang)

extra-a-test-1-gpu-large-amd${{ ((inputs.rocm_version && (inputs.rocm_version != 'rocm724') && format('-{0}', inputs.rocm_version)) || '') }} (linux-${{ (inputs.runner_arch || 'mi300') }}-1gpu-sglang)

extra-a-test-2-gpu-large-amd${{ ((inputs.rocm_version && (inputs.rocm_version != 'rocm724') && format('-{0}', inputs.rocm_version)) || '') }} (linux-${{ (inputs.runner_arch || 'mi300') }}-2gpu-sglang)

pr-test-amd-extra-finish

PR Test (XPU) on: pull_request 3

check-changes

pr-gate / pr-gate

stage-a-test-1-gpu-xpu

wait-for-stage-a

stage-b-test-1-gpu-xpu

finish

PR Test (MUSA) on: pull_request 1

check-changes

pr-gate / pr-gate

multimodal-gen-test-1-gpu-musa

multimodal-gen-test-2-gpu-musa

multimodal-gen-layer-unit-test-musa

sgl-kernel-unit-test-musa

pr-test-musa-finish

PR Test (MLX) on: pull_request 5

check-changes

pr-gate / pr-gate

stage-a-unit-test-mlx

stage-b-e2e-test-mlx

pr-test-mlx-finish

Auto Label PRs on: pull_request_target 1

label

PR States on: pull_request_target

update-pr-body

PR Test (Arm64) on: pull_request 2

check-changes

pr-gate / pr-gate

build-test

PR Test ROCm 7.2 (AMD) on: pull_request 7

call-gate / pr-gate

call-pr-test-amd-extra-rocm720

check-changes

sgl-kernel-unit-test-amd-rocm720 (linux-${{ (inputs.runner_arch || 'mi300') }}-1gpu-sglang)

sgl-kernel-unit-test-2-gpu-amd-rocm720 (linux-${{ (inputs.runner_arch || 'mi300') }}-2gpu-sglang)

stage-a-test-1-gpu-small-amd-rocm720 (linux-mi300-1gpu-sglang)

jit-kernel-unit-test-amd-rocm720 (linux-mi300-1gpu-sglang)

jit-kernel-benchmark-test-amd-rocm720 (linux-mi300-1gpu-sglang)

wait-for-stage-a-amd-rocm720

multimodal-gen-test-1-gpu-amd-rocm720 (linux-${{ (inputs.runner_arch || 'mi300') }}-1gpu-sglang, ${{ matrix.part }})

multimodal-gen-test-2-gpu-amd-rocm720 (linux-${{ (inputs.runner_arch || 'mi300') }}-2gpu-sglang, ${{ matrix.part }})

stage-b-test-1-gpu-small-amd-rocm720 (linux-${{ (inputs.runner_arch || 'mi300') }}-1gpu-sglang, ${{ matrix.part }})

stage-b-test-1-gpu-small-amd-nondeterministic-rocm720 (linux-${{ (inputs.runner_arch || 'mi300') }}-1gpu-sglang)

stage-b-test-1-gpu-small-amd-mi35x-rocm720

stage-b-test-1-gpu-large-amd-rocm720 (linux-${{ (inputs.runner_arch || 'mi300') }}-1gpu-sglang, ${{ matrix.part }})

stage-b-test-2-gpu-large-amd-rocm720 (linux-${{ (inputs.runner_arch || 'mi300') }}-2gpu-sglang, ${{ matrix.part }})

wait-for-stage-b-amd-rocm720

stage-b-test-large-8-gpu-mi35x-disaggregation-amd-rocm720

stage-c-test-4-gpu-amd-rocm720 (linux-${{ (inputs.runner_arch || 'mi300') }}-4gpu-sglang, ${{ matrix.part }})

stage-c-test-large-8-gpu-amd-rocm720 (linux-${{ (inputs.runner_arch || 'mi300') }}-8gpu-sglang, ${{ matrix.part }})

stage-c-test-large-8-gpu-amd-mi35x-rocm720

stage-c-dsv4-flash-fp4-fp8-amd-mi35x-rocm720

stage-c-dsv4-pro-fp4-amd-mi35x-rocm720

pr-test-amd-rocm720-finish

PR Test (NPU) on: pull_request 61

check-changes

set-image-config

pr-gate / pr-gate

base-a-test-1-npu-a2 / run (0)

base-b-test-1-npu-a3 / run (0)

base-b-test-2-npu-a3 / run (0)

base-b-test-4-npu-a3 / run (0)

base-b-test-4-npu-a3 / run (1)

base-b-test-8-npu-a3 / run (0)

base-b-test-16-npu-a3 / run (0)

multimodal-gen-test-1-npu-a3

multimodal-gen-test-2-npu-a3

base-c-test-acc-2-npu-a3 / base-c-test-acc-2-npu-a3

base-c-test-acc-4-npu-a3 / base-c-test-acc-4-npu-a3

base-c-test-acc-8-npu-a3 / base-c-test-acc-8-npu-a3

base-c-test-acc-16-npu-a3 / base-c-test-acc-16-npu-a3

base-c-test-perf-2-npu-a3

base-c-test-perf-4-npu-a3 / base-c-test-perf-4-npu-a3

base-c-test-perf-8-npu-a3 / base-c-test-perf-8-npu-a3

base-c-test-perf-16-npu-a3 / base-c-test-perf-16-npu-a3

pr-test-npu-finish

PR Test Base on: pull_request 57

check-changes / check-changes

call-pr-test-extra

call-gate / pr-gate

wait-for-base-a

sgl-kernel-build-wheels

sgl-kernel-build-wheels-arm

rust-ext-build / Restore Rust Ext

rust-ext-build / Build Rust Ext

wait-for-base-b

call-sgl-kernel-tests

call-jit-kernel-tests / jit-kernel-unit-test

call-jit-kernel-tests / jit-kernel-multigpu-unit-test

call-jit-kernel-tests / jit-kernel-benchmark-test

call-jit-kernel-tests / jit-kernel-b200-test

base-a-test-1-gpu-small / base-a-test-1-gpu-small (0)

base-a-test-cpu / base-a-test-cpu (0)

base-a-test-cpu / base-a-test-cpu (1)

base-a-test-cpu / base-a-test-cpu (2)

base-a-test-cpu / base-a-test-cpu (3)

base-a-test-cpu / base-a-test-cpu (4)

base-a-test-cpu / base-a-test-cpu (5)

base-a-test-cpu / base-a-test-cpu (6)

base-a-test-cpu / base-a-test-cpu (7)

base-a-test-cpu / base-a-test-cpu (8)

base-a-test-cpu / base-a-test-cpu (9)

base-a-test-cpu / base-a-test-cpu (10)

base-b-test-1-gpu-small / base-b-test-1-gpu-small (0)

base-b-test-1-gpu-small / base-b-test-1-gpu-small (1)

base-b-test-1-gpu-small / base-b-test-1-gpu-small (2)

base-b-test-1-gpu-small / base-b-test-1-gpu-small (3)

base-b-test-1-gpu-small / base-b-test-1-gpu-small (4)

base-b-test-1-gpu-small / base-b-test-1-gpu-small (5)

base-b-test-1-gpu-small / base-b-test-1-gpu-small (6)

base-b-test-1-gpu-small / base-b-test-1-gpu-small (7)

base-b-test-1-gpu-large / base-b-test-1-gpu-large (0)

base-b-test-1-gpu-large / base-b-test-1-gpu-large (1)

base-b-test-1-gpu-large / base-b-test-1-gpu-large (2)

base-b-test-1-gpu-large / base-b-test-1-gpu-large (3)

base-b-test-1-gpu-large / base-b-test-1-gpu-large (4)

base-b-test-1-gpu-large / base-b-test-1-gpu-large (5)

base-b-test-1-gpu-large / base-b-test-1-gpu-large (6)

base-b-test-1-gpu-large / base-b-test-1-gpu-large (7)

base-b-test-2-gpu-large / base-b-test-2-gpu-large (0)

base-b-test-2-gpu-large / base-b-test-2-gpu-large (1)

base-b-test-2-gpu-large / base-b-test-2-gpu-large (2)

base-b-test-2-gpu-large / base-b-test-2-gpu-large (3)

base-b-test-2-gpu-large / base-b-test-2-gpu-large (4)

base-b-test-4-gpu-b200 / base-b-test-4-gpu-b200 (0)

call-multimodal-gen-tests

base-c-test-4-gpu-h100 / base-c-test-4-gpu-h100 (0)

base-c-test-4-gpu-h100 / base-c-test-4-gpu-h100 (1)

base-c-test-4-gpu-h100 / base-c-test-4-gpu-h100 (2)

base-c-test-8-gpu-h200 / base-c-test-8-gpu-h200 (0)

base-c-test-8-gpu-h200 / base-c-test-8-gpu-h200 (1)

base-c-test-8-gpu-h200 / base-c-test-8-gpu-h200 (2)

base-c-test-8-gpu-h20 / base-c-test-8-gpu-h20 (0)

base-c-test-8-gpu-h20 / base-c-test-8-gpu-h20 (1)

base-c-test-8-gpu-h20 / base-c-test-8-gpu-h20 (2)

base-c-test-4-gpu-b200 / base-c-test-4-gpu-b200 (0)

base-c-test-4-gpu-b200 / base-c-test-4-gpu-b200 (1)

base-c-test-4-gpu-b200 / base-c-test-4-gpu-b200 (2)

base-c-test-4-gpu-gb300 / base-c-test-4-gpu-gb300 (0)

base-c-test-4-gpu-gb300 / base-c-test-4-gpu-gb300 (1)

base-c-test-8-gpu-b300 / base-c-test-8-gpu-b300 (0)

pr-test-finish

notify-pr-states

Annotations

1 warning

Complete job

Node.js 20 is deprecated. The following actions target Node.js 20 but are being forced to run on Node.js 24: actions/checkout@v4, dorny/paths-filter@v3. For more information see: https://github.blog/changelog/2025-09-19-deprecation-of-node-20-on-github-actions-runners/

gate

succeeded Aug 22, 2026 in 10s

Loading

Re-running jobs...

Footer

© 2026 GitHub, Inc.

Footer navigation

Terms
Privacy
Security
Status
Community
Docs
Contact

You can’t perform that action at this time.