infinigence / Semi-PDLinks
A prefill & decode disaggregated LLM serving framework with shared GPU memory and fine-grained compute isolation.
☆123Updated 2 weeks ago
Alternatives and similar repositories for Semi-PD
Users that are interested in Semi-PD are comparing it to the libraries listed below
Sorting:
- A lightweight design for computation-communication overlap.☆209Updated 2 weeks ago
- High performance Transformer implementation in C++.☆148Updated 11 months ago
- NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer☆152Updated 3 months ago
- DLSlime: Flexible & Efficient Heterogeneous Transfer Toolkit☆91Updated last week
- ☆92Updated 9 months ago
- DeepSeek-V3/R1 inference performance simulator☆176Updated 9 months ago
- FlagCX is a scalable and adaptive cross-chip communication library.☆166Updated this week
- Stateful LLM Serving☆93Updated 10 months ago
- Aims to implement dual-port and multi-qp solutions in deepEP ibrc transport☆73Updated 8 months ago
- DeepXTrace is a lightweight tool for precisely diagnosing slow ranks in DeepEP-based environments.☆85Updated 3 weeks ago
- PyTorch distributed training acceleration framework☆54Updated 5 months ago
- Paella: Low-latency Model Serving with Virtualized GPU Scheduling☆66Updated last year
- Compare different hardware platforms via the Roofline Model for LLM inference tasks.☆120Updated last year
- FlexFlow Serve: Low-Latency, High-Performance LLM Serving☆66Updated 3 months ago
- ☆81Updated 2 months ago
- Artifact of OSDI '24 paper, ”Llumnix: Dynamic Scheduling for Large Language Model Serving“☆64Updated last year
- ☆104Updated last year
- ☆152Updated last year
- ☆338Updated last week
- Summary of the Specs of Commonly Used GPUs for Training and Inference of LLM☆71Updated 5 months ago
- ☆112Updated 7 months ago
- Tile-based language built for AI computation across all scales☆115Updated 3 weeks ago
- Offline optimization of your disaggregated Dynamo graph☆146Updated this week
- ☆47Updated last year
- Open ABI and FFI for Machine Learning Systems☆293Updated this week
- Venus Collective Communication Library, supported by SII and Infrawaves.☆132Updated this week
- DeeperGEMM: crazy optimized version☆74Updated 8 months ago
- ☆96Updated 9 months ago
- SpotServe: Serving Generative Large Language Models on Preemptible Instances☆134Updated last year
- A standalone GEMM kernel for fp16 activation and quantized weight, extracted from FasterTransformer☆96Updated 4 months ago