Skip to content
Navigation Menu
Sign in
Products
Platform
GitHub
Repository hosting, code review, and collaboration
Copilot
AI coding agent for tasks, projects, and automated work
Build & Ship
Actions
CI/CD and automation
Code Quality
Quality analysis, fixes, and enforcement
Codespaces
Cloud development environments
Secure
Code Security
Vulnerability detection and remediation
Secret Protection
Secret scanning and leak prevention
All Features
Developers
Learn
Documentation
Training
Certifications
Build
Build with Copilot
Customize Copilot
Integrations
Explore
Trending repos
GitHub Sponsors
Developer articles
Changelog
Changelog
Your source for the latest features, improvements, and fixes across GitHub products, all in one place.
Businesses
By Stage
Enterprises
Team
Startups
Nonprofits
By Industry
Healthcare
Financial services
Manufacturing
Government
See all industries
Evaluate GitHub
Why GitHub
Customer Stories
Business insights
Trust Center
Add-ons
GitHub Advanced Security
Enterprise-grade code security
Copilot for business
Enterprise-grade AI features
Premium Support
Enterprise-grade 24/7 support
Resources
Learn & Connect
Documentation
Resource center
Blog
Events & webinars
Community Forum
Customer Stories
Integrations
Support
Customer support
Training
GitHub Status
Company
About
Partners
Changelog
GitHub Universe
GitHub Universe
The event for developers, by developers — talks, workshops, and what’s next for GitHub.
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 }}
certik
lfortran
Repository navigation
Code
Pull requests
1
(1)
Actions
Projects
Security and quality
Insights
More
items
Formal on NVIDIA GPUs: integrated --gpu=cuda stack + hardware runbook
- #10
#10
Draft
certik
wants to merge 29 commits into
formal-
cuda-
nvidia-
base
certik/lfortran:formal-cuda-nvidia-base
from
test/
formal-
cuda-
nvidia
certik/lfortran:test/formal-cuda-nvidia
Copy head branch name to clipboard
Conversation
Commits
29
(29)
Checks
Files changed
Draft
Formal on NVIDIA GPUs: integrated --gpu=cuda stack + hardware runbook
#10
certik
wants to merge 29 commits into
formal-
cuda-
nvidia-
base
certik/lfortran:formal-cuda-nvidia-base
from
test/
formal-
cuda-
nvidia
certik/lfortran:test/formal-cuda-nvidia
Copy head branch name to clipboard
Commits
Commits on Oct 18, 2026
fix: regenerate CUDA sources per translation unit
certik
committed
e4bb7dc
View commit details
Copy full SHA for e4bb7dc
Browse repository at this point
fix: frame CUDA archive metadata
certik
committed
91ae734
View commit details
Copy full SHA for 91ae734
Browse repository at this point
fix: keep native CUDA registrar local
certik
committed
161e7a2
View commit details
Copy full SHA for 161e7a2
Browse repository at this point
fix: resolve CUDA archive libraries before source collection
certik
committed
ac9c757
View commit details
Copy full SHA for ac9c757
Browse repository at this point
Commits on Oct 19, 2026
fix: cover CUDA library selection helpers
certik
committed
fe16b8a
View commit details
Copy full SHA for fe16b8a
Browse repository at this point
fix: compile CUDA device code for the target GPU architecture
Show description for b2f389b
certik
committed
b2f389b
View commit details
Copy full SHA for b2f389b
Browse repository at this point
fix: make formal CUDA fixtures portable
certik
committed
33ce46e
View commit details
Copy full SHA for 33ce46e
Browse repository at this point
fix: give each CUDA translation unit a uniquely named kernel registrar
Show description for 5430573
certik
committed
5430573
View commit details
Copy full SHA for 5430573
Browse repository at this point
fix: skip a GPU launch with no threads in the CUDA runtime
Show description for 6a5d5a1
certik
committed
6a5d5a1
View commit details
Copy full SHA for 6a5d5a1
Browse repository at this point
test: run the GPU tests that pass on an NVIDIA GPU with the cuda backend
Show description for 24efaca
certik
committed
24efaca
View commit details
Copy full SHA for 24efaca
Browse repository at this point
fix: give each GPU thread its own copy of per-iteration arrays
Show description for f1bbb3a
certik
committed
f1bbb3a
View commit details
Copy full SHA for f1bbb3a
Browse repository at this point
Merge branch 'pr-15546' into test/formal-cuda-nvidia
certik
committed
5033cd4
View commit details
Copy full SHA for 5033cd4
Browse repository at this point
Merge branch 'pr-15559' into test/formal-cuda-nvidia
Show description for a1cc5c7
certik
committed
a1cc5c7
View commit details
Copy full SHA for a1cc5c7
Browse repository at this point
Merge branch 'pr-15560' into test/formal-cuda-nvidia
Show description for 4f1f3fb
certik
committed
4f1f3fb
View commit details
Copy full SHA for 4f1f3fb
Browse repository at this point
Merge branch 'pr-15562' into test/formal-cuda-nvidia
certik
committed
09a1155
View commit details
Copy full SHA for 09a1155
Browse repository at this point
Merge branch 'pr-15563' into test/formal-cuda-nvidia
Show description for 2c09997
certik
committed
2c09997
View commit details
Copy full SHA for 2c09997
Browse repository at this point
test: restore cuda labels from #15562/#15563 dropped while merging
Show description for e569d5a
certik
committed
e569d5a
View commit details
Copy full SHA for e569d5a
Browse repository at this point
Merge branch 'pr-15569' into test/formal-cuda-nvidia
Show description for 386048e
certik
committed
386048e
View commit details
Copy full SHA for 386048e
Browse repository at this point
fix: declare only the variables a GPU kernel scope refers to
Show description for 8d5b1ed
certik
committed
8d5b1ed
View commit details
Copy full SHA for 8d5b1ed
Browse repository at this point
fix: evaluate each operand of a defined operator once
Show description for 0dadf0c
certik
committed
0dadf0c
View commit details
Copy full SHA for 0dadf0c
Browse repository at this point
fix: forget tracked counters across calls in GPU array privatization
Show description for f13ef24
certik
committed
f13ef24
View commit details
Copy full SHA for f13ef24
Browse repository at this point
Merge branch 'pr-15569' into test/formal-cuda-nvidia
certik
committed
d5c48f1
View commit details
Copy full SHA for d5c48f1
Browse repository at this point
Merge branch 'pr-15570' into test/formal-cuda-nvidia
Show description for e51d12a
certik
committed
e51d12a
View commit details
Copy full SHA for e51d12a
Browse repository at this point
Merge branch 'pr-15578' into test/formal-cuda-nvidia
certik
committed
b69450c
View commit details
Copy full SHA for b69450c
Browse repository at this point
perf(gpu): reuse CUDA scratch buffers across launches
Show description for 9021c8b
certik
committed
9021c8b
View commit details
Copy full SHA for 9021c8b
Browse repository at this point
Merge branch 'pr-15580' into test/formal-cuda-nvidia
Show description for 65b945d
certik
committed
65b945d
View commit details
Copy full SHA for 65b945d
Browse repository at this point
test: regenerate gpu_metal_474 CUDA reference for the combined stack
Show description for 97301a0
certik
committed
97301a0
View commit details
Copy full SHA for 97301a0
Browse repository at this point
ci: document and script validating --gpu=cuda on NVIDIA hardware with Formal
Show description for 3ba8e18
certik
committed
3ba8e18
View commit details
Copy full SHA for 3ba8e18
Browse repository at this point
ci: check that every do concurrent in Formal is offloaded
Show description for 93826d8
certik
committed
93826d8
View commit details
Copy full SHA for 93826d8
Browse repository at this point
You can’t perform that action at this time.