repoGitHubTrust 82 · PrimaryPublished 1mo agoLive · 26d ago
uccl-project/uccl
UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g., GPU-driven)
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 51%WattGPU: Predicting Inference Power and Latency on Unseen GPUs and LLMs →
- PossiblePossibly related (embedding) · 48%Anyone using TensTorrent gpus for your local ai? What's been your experience? →
- PossiblePossibly related (embedding) · 47%GPU Parallelization Strategies for Forward and Backward Propagation in Shallow Neural Networks: A CUDA-Based Comparative Study →
- PossiblePossibly related (embedding) · 45%Build real agentic apps using CUGA: two dozen working examples on a lightweight harness →
- PossiblePossibly related (embedding) · 54%Top Cost-Effective Enterprise GPU Cloud Platforms for AI Workloads with H100–GB200, Elastic Scaling and Pay-as-You-Go Compute - Scott Coop →
- PossiblePossibly related (embedding) · 50%Learning FlashAttention the Hard Way. Part 1: The Algebraic Foundation [D] →
- PossiblePossibly related (embedding) · 46%Recent llama.cpp updates for SYCL/Intel →
- PossiblePossibly related (embedding) · 47%Show HN: Computable – Buy, sell, and redeem GPU for the exact weeks you want →
Implements
Covers
Covers (incoming)
newsTop Cost-Effective Enterprise GPU Cloud Platforms for AI Workloads with H100–GB200, Elastic Scaling and Pay-as-You-Go Compute - Scott CoopnewsLearning FlashAttention the Hard Way. Part 1: The Algebraic Foundation [D]newsRecent llama.cpp updates for SYCL/IntelnewsShow HN: Computable – Buy, sell, and redeem GPU for the exact weeks you wantnewsPSA: DO NOT use Intel consumer platforms for multi-GPU setups
Related across the graph
newsLearning FlashAttention the Hard Way. Part 1: The Algebraic Foundation [D]newsRecent llama.cpp updates for SYCL/IntelnewsBuild real agentic apps using CUGA: two dozen working examples on a lightweight harnessnewsPSA: DO NOT use Intel consumer platforms for multi-GPU setupspaperWattGPU: Predicting Inference Power and Latency on Unseen GPUs and LLMspaperGPU Parallelization Strategies for Forward and Backward Propagation in Shallow Neural Networks: A CUDA-Based Comparative StudynewsShow HN: Computable – Buy, sell, and redeem GPU for the exact weeks you wantnewsAnyone using TensTorrent gpus for your local ai? What's been your experience?newsTop Cost-Effective Enterprise GPU Cloud Platforms for AI Workloads with H100–GB200, Elastic Scaling and Pay-as-You-Go Compute - Scott Coop
