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
  Accelerator
  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

Search code, repositories, users, issues, pull requests...

Search

Clear

Search syntax tips

Provide feedback

We read every piece of feedback, and take your input very seriously.

Include my email address so I can be contacted

Saved searches

Use saved searches to filter your results more quickly

Name

Query

To see all available qualifiers, see our documentation.

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

ggml-org / llama.cpp Public

Notifications You must be signed in to change notification settings
Fork 16k
Star 99.8k

Code
Issues 485
Pull requests 815
Discussions
Actions
Projects
Wiki
Security 13
Insights

Additional navigation options

Code
Issues
Pull requests
Discussions
Actions
Projects
Wiki
Security
Insights

ggml: backend-agnostic tensor parallelism#19378

Open

JohannesGaessler wants to merge 45 commits intoggml-org:masterggml-org/llama.cpp:masterfrom

JohannesGaessler:ggml-meta-backend-8JohannesGaessler/llama.cpp:ggml-meta-backend-8Copy head branch name to clipboard

Conversation Commits45 (45)Checks Files changed

Open

ggml: backend-agnostic tensor parallelism#19378
JohannesGaessler wants to merge 45 commits intoggml-org:masterfrom
JohannesGaessler:ggml-meta-backend-8

Commits

Commits on Mar 25, 2026

ggml: backend-agnostic tensor parallelism
JohannesGaessler
committed
support for GPT-OSS, Qwen 3 MoE
JohannesGaessler
committed
partial Vulkan fix
JohannesGaessler
committed
add support for 4/8 GPUs
JohannesGaessler
committed
unconditional peer access
JohannesGaessler
committed
re-use buffers + ggml contexts
JohannesGaessler
committed
fix output pattern
JohannesGaessler
committed
NCCL support
JohannesGaessler
committed
GGML: HIP: add RCCL support

IMbackK
authored and
JohannesGaessler
committed
Remove shfl and AllReduce from backend interface
JohannesGaessler
committed
move allocation workaround out of ggml-alloc.c
JohannesGaessler
committed
2d tensor set/get support
JohannesGaessler
committed
Fix the seg fault without NCCL

gaugarg-nv
authored and
JohannesGaessler
committed
Apply suggestion from @JohannesGaessler
JohannesGaessler
committed
support for tensor dims % n_devs != 0
JohannesGaessler
committed
fix view_offs scaling
JohannesGaessler
committed
arbitrary num. of GPUs/tensor split
JohannesGaessler
committed
fix compilation
JohannesGaessler
committed
better granularity estimate
JohannesGaessler
committed
Support device-specific host buffer types if all underlying backends expose the same type. This allows using pinned memory instead of pageable memory for CUDA.

gaugarg-nv
authored and
JohannesGaessler
committed
partial Qwen 3 Next support
JohannesGaessler
committed
Fix qwen3 30b (#8)

gaugarg-nv
authored and
JohannesGaessler
committed
Fix crashes due to KV cache serialization (#9)

gaugarg-nv
authored and
JohannesGaessler
committed
metal : fix build (#7)

ggerganov
authored and
JohannesGaessler
committed
static memory allocations, fix usage count
JohannesGaessler
committed
fix tensor granularity
JohannesGaessler
committed
more even memory distribution
JohannesGaessler
committed
use BF16 for allreduce
JohannesGaessler
committed
rebase fixup
JohannesGaessler
committed
better error message for unsupported architectures
JohannesGaessler
committed
Fix device mismatch during scatter of allReduce. (#11)

gaugarg-nv
authored and
JohannesGaessler
committed
Enable the previous allreduce implementation. It is better in both perf and stability (#12)

gaugarg-nv
authored and
JohannesGaessler
committed
delay AllReduce for Moe for less I/O
JohannesGaessler
committed
build : clean-up compile warnings

ggerganov
authored and
JohannesGaessler
committed
backend : move most of the meta backend API to ggml-backend-impl.h

ggerganov
authored and
JohannesGaessler
committed
cont : hide unused public API in the implementation

ggerganov
authored and
JohannesGaessler
committed
llama : use llama_device + remove ggml_backend_dev_is_meta()

ggerganov
authored and
JohannesGaessler
committed
ggml-backend : remove unused alloc include

ggerganov
authored and
JohannesGaessler
committed
minor : remove regex include

ggerganov
authored and
JohannesGaessler
committed
ggml : introduce ggml-ext.h for staging new APIs

ggerganov
authored and
JohannesGaessler
committed
rebase fixup
JohannesGaessler
committed
fix tests
JohannesGaessler
committed

Commits on Mar 26, 2026

llama : more robust logic for determining Meta devices (#16)

ggerganov
and
JohannesGaessler
authored
disable roundtrip for meta backend
JohannesGaessler
committed
fix arch selection
JohannesGaessler
committed

Footer

© 2026 GitHub, Inc.

Footer navigation

Terms
Privacy
Security
Status
Community
Docs
Contact

You can’t perform that action at this time.