OrbitInfra

All

6 repositories

OrbiTE
Public
OrbiTE is a transfer engine for LLM inference.
rust cuda gpudirect-rdma
rust cuda gpudirect-rdma rdma nvlink io-uring kv-cache ai-infra llm-inference gpudirect-storage
Rust
•
Apache License 2.0
•1•2•0•1•Updated Nov 13, 2026Nov 13, 2026
OrbitKV
Public
OrbitKV is a KV cache for vLLM and SGLang.
rust attention memory-management
rust attention memory-management kv-cache llm-serving vllm llm-inference sglang prefix-caching orbitkv
Rust
•
Apache License 2.0
•19•282•1•1•Updated Nov 13, 2026Nov 13, 2026
DeepInfer
Public
DeepInfer: High-performance, large-scale inference framework for DeepSeek models
Python
•
Apache License 2.0
•3.2k•0•0•0•Updated Nov 8, 2026Nov 8, 2026
Mooncake
Public
Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.
C++
•
Apache License 2.0
•1.5k•0•0•0•Updated Nov 8, 2026Nov 8, 2026
DeepEP
Public
DeepEP: an efficient expert-parallel communication library
Cuda
•
MIT License
•1.6k•0•0•0•Updated Oct 21, 2026Oct 21, 2026
DeepGEMM
Public
DeepGEMM: clean and efficient BLAS kernel library on GPU
Cuda
•
MIT License
•1.5k•0•0•0•Updated Oct 20, 2026Oct 20, 2026

ProTip! When viewing an organization's repositories, you can use the props. filter to filter by custom property.