Skip to content
Navigation Menu
Sign in
Appearance settings
Platform
AI CODE CREATION
GitHub Copilot
Write better code with AI
GitHub Copilot app
Direct agents from issue to merge
MCP Registry
Integrate external tools
DEVELOPER WORKFLOWS
Actions
Automate any workflow
Codespaces
Instant dev environments
Issues
Plan and track work
Code Review
Manage code changes
Code Quality
Enforce quality at merge
APPLICATION SECURITY
GitHub Advanced Security
Find and fix vulnerabilities
Code security
Secure your code as you build
Secret protection
Stop leaks before they start
EXPLORE
Why GitHub
Documentation
Blog
Changelog
Marketplace
View all features
Solutions
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
Resources
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
Open Source
COMMUNITY
GitHub Sponsors
Fund open source developers
PROGRAMS
Security Lab
Maintainer Community
GitHub Stars
Archive Program
REPOSITORIES
Topics
Trending
Collections
Enterprise
ENTERPRISE SOLUTIONS
Enterprise platform
AI-powered developer platform
AVAILABLE ADD-ONS
GitHub Advanced Security
Enterprise-grade security features
Copilot for Business
Enterprise-grade AI features
Premium Support
Enterprise-grade 24/7 support
Pricing
Search
/
Sign in
Sign up
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
{{ message }}
NVIDIA
/
TensorRT-LLM
Public
Notifications
You must be signed in to change notification settings
Fork
2.8k
Star
14.7k
Code
Issues
606
Pull requests
903
Discussions
Actions
Projects
Security and quality
0
Insights
Additional navigation options
Code
Issues
Pull requests
Discussions
Actions
Projects
Security and quality
Insights
[None][feat] Rubin fp4 mla refactor(P1)
- #18478
#18478
Merged
Tracin
merged 21 commits into
NVIDIA:main
NVIDIA/TensorRT-LLM:main
from
Tracin:feat/rubin-fp4-mla-core-main
Tracin/TensorRT-LLM:feat/rubin-fp4-mla-core-main
Copy head branch name to clipboard
Sep 22, 2026
Conversation
Commits
21
(21)
Checks
Files changed
Merged
[None][feat] Rubin fp4 mla refactor(P1)
#18478
Tracin
merged 21 commits into
NVIDIA:main
NVIDIA/TensorRT-LLM:main
from
Tracin:feat/rubin-fp4-mla-core-main
Tracin/TensorRT-LLM:feat/rubin-fp4-mla-core-main
Copy head branch name to clipboard
Commits
Commits on Sep 22, 2026
[None][feat] add FP4 MLA attention backend on Rubin
Show description for 2c71ac1
Tracin
committed
2c71ac1
View commit details
Copy full SHA for 2c71ac1
Browse repository at this point
[None][feat] integrate FP4 MLA KV cache manager V2
Show description for d199b53
Tracin
committed
d199b53
View commit details
Copy full SHA for d199b53
Browse repository at this point
[None][test] add FP4 MLA core coverage
Show description for 342559e
Tracin
committed
342559e
View commit details
Copy full SHA for 342559e
Browse repository at this point
Migrate FP4 MLA kernels to public CuTeDSL APIs
Show description for 777a3ff
root
authored and
Tracin
committed
777a3ff
View commit details
Copy full SHA for 777a3ff
Browse repository at this point
[None][fix] resolve FP4 MLA pre-commit failures
Show description for 1f4232f
Tracin
committed
1f4232f
View commit details
Copy full SHA for 1f4232f
Browse repository at this point
[None][fix] gate FP4 MLA fallback during FMHA registration
Show description for 76bd06c
Tracin
committed
76bd06c
View commit details
Copy full SHA for 76bd06c
Browse repository at this point
[None][fix] guard optional FP4 MLA metadata hooks
Show description for 85a02e4
Tracin
committed
85a02e4
View commit details
Copy full SHA for 85a02e4
Browse repository at this point
[None][fix] preserve executor contracts with FP4 MLA routing
Show description for b10e9c6
Tracin
committed
b10e9c6
View commit details
Copy full SHA for b10e9c6
Browse repository at this point
[None][fix] preserve attention metadata compatibility for FP4 MLA
Show description for 4c37a41
Tracin
committed
4c37a41
View commit details
Copy full SHA for 4c37a41
Browse repository at this point
[None][fix] support large KV lengths in FP4 MLA kernels
Show description for c817353
Tracin
committed
c817353
View commit details
Copy full SHA for c817353
Browse repository at this point
fix: address FP4 MLA core review feedback
Show description for adebf19
Tracin
committed
adebf19
View commit details
Copy full SHA for adebf19
Browse repository at this point
fix: accept integer quant modes in FP4 MLA routing
Show description for 1f477dd
Tracin
committed
1f477dd
View commit details
Copy full SHA for 1f477dd
Browse repository at this point
fix: preserve quant mode wrappers during config copying
Show description for cc98243
Tracin
committed
cc98243
View commit details
Copy full SHA for cc98243
Browse repository at this point
test: align MLA cache reuse mocks with model config
Show description for 941216b
Tracin
committed
941216b
View commit details
Copy full SHA for 941216b
Browse repository at this point
refactor: address FP4 MLA capability and metadata review
Show description for f756d4c
Tracin
committed
f756d4c
View commit details
Copy full SHA for f756d4c
Browse repository at this point
refactor: streamline FP4 MLA runtime helpers
Show description for 17fe505
Tracin
committed
17fe505
View commit details
Copy full SHA for 17fe505
Browse repository at this point
test: preserve cache hits in FMHA sanity check coverage
Show description for 711a199
Tracin
committed
711a199
View commit details
Copy full SHA for 711a199
Browse repository at this point
refactor: simplify FP4 MLA cache policy and RoPE sizing
Show description for 7241ddf
Tracin
committed
7241ddf
View commit details
Copy full SHA for 7241ddf
Browse repository at this point
refactor: integrate upstream FP4 MLA cache policy and helpers
Show description for 70f6f2e
Tracin
committed
70f6f2e
View commit details
Copy full SHA for 70f6f2e
Browse repository at this point
bench: add FP4 MLA decode microbenchmark
Show description for 80e9543
Tracin
committed
80e9543
View commit details
Copy full SHA for 80e9543
Browse repository at this point
fix: isolate dense FP4 MLA dispatch and remove stale state
Show description for b0d8b98
Tracin
committed
b0d8b98
View commit details
Copy full SHA for b0d8b98
Browse repository at this point
You can’t perform that action at this time.
