A prefill & decode disaggregated LLM serving framework with shared GPU memory and fine-grained compute isolation.
☆130Dec 25, 2025Updated 7 months ago
Alternatives and similar repositories for Semi-PD
Users that are interested in Semi-PD are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- Repo for SpecEE: Accelerating Large Language Model Inference with Speculative Early Exiting (ISCA25)☆75Apr 25, 2025Updated last year
- A lightweight design for computation-communication overlap.☆243Jan 20, 2026Updated 6 months ago
- ☆46Oct 15, 2025Updated 10 months ago
- High-performance distributed data shuffling (all-to-all) library for MoE training and inference☆123Mar 7, 2026Updated 5 months ago
- KV cache store for distributed LLM inference☆430Nov 13, 2025Updated 9 months ago
- Proton VPN Special Offer - Get 70% off • AdSpecial partner offer. Trusted by over 100 million users worldwide. Tested, Approved and Recommended by Experts.
- Efficient and easy multi-instance LLM serving☆564Mar 12, 2026Updated 5 months ago
- ☆65Apr 26, 2025Updated last year
- Materials for learning SGLang☆875Jan 5, 2026Updated 7 months ago
- ☆382Jan 28, 2026Updated 6 months ago
- DeeperGEMM: crazy optimized version☆86May 5, 2025Updated last year
- NVIDIA Inference Xfer Library (NIXL)☆1,195Updated this week
- A fast communication-overlapping library for tensor/expert parallelism on GPUs.☆1,352Aug 28, 2025Updated 11 months ago
- An experimental communicating attention kernel based on DeepEP.☆34Jul 29, 2025Updated last year
- DuoDecoding: Hardware-aware Heterogeneous Speculative Decoding with Dynamic Multi-Sequence Drafting☆19Mar 4, 2025Updated last year
- Managed Database hosting by DigitalOcean • AdPostgreSQL, MySQL, MongoDB, Kafka, Valkey, and OpenSearch available. Automatically scale up storage and focus on building your apps.
- Perplexity GPU Kernels☆601Nov 7, 2025Updated 9 months ago
- DeepSeek-V3/R1 inference performance simulator☆194Mar 27, 2025Updated last year
- a size profiler for cuda binary☆71Jan 15, 2026Updated 7 months ago
- ☆91Oct 17, 2025Updated 10 months ago
- Disaggregated serving system for Large Language Models (LLMs).☆830Apr 6, 2025Updated last year
- Boosting GPU utilization for LLM serving via dynamic spatial-temporal prefill & decode orchestration☆52Jan 8, 2026Updated 7 months ago
- Distributed Compiler and Optimized Parallel Kernels☆1,518Updated this week
- ☆54May 19, 2025Updated last year
- Dynamic Memory Management for Serving LLMs without PagedAttention☆513Jul 17, 2026Updated last month
- Proton VPN Special Offer - Get 70% off • AdSpecial partner offer. Trusted by over 100 million users worldwide. Tested, Approved and Recommended by Experts.
- Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.☆6,313Updated this week
- ☆75Mar 26, 2025Updated last year
- ☆135Nov 11, 2024Updated last year
- ☆107Sep 9, 2024Updated last year
- Accurate, large-scale, and extensible simulator for LLM inference Systems☆666Jul 25, 2025Updated last year
- ☆65Updated this week
- NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer☆198Feb 11, 2026Updated 6 months ago
- [NeurIPS 2025] ClusterFusion: Expanding Operator Fusion Scope for LLM Inference via Cluster-Level Collective Primitive☆75Aug 8, 2026Updated last week
- RTP-LLM: Alibaba's high-performance LLM inference engine for diverse applications.☆1,309Updated this week
- Managed Database hosting by DigitalOcean • AdPostgreSQL, MySQL, MongoDB, Kafka, Valkey, and OpenSearch available. Automatically scale up storage and focus on building your apps.
- Open ABI and FFI for Machine Learning Systems☆446Updated this week
- ☆42Oct 12, 2025Updated 10 months ago
- Train speculative decoding models effortlessly and port them smoothly to SGLang serving.☆1,090Updated this week
- MSCCL++: A GPU-driven communication stack for scalable AI applications☆549Updated this week
- ☆320Updated this week
- Medusa: Accelerating Serverless LLM Inference with Materialization [ASPLOS'25]☆12Nov 8, 2024Updated last year
- Injecting Adrenaline into LLM Serving: Boosting Resource Utilization and Throughput via Attention Disaggregation☆42Updated this week