torchcomms: a modern PyTorch communications API
☆380Jul 23, 2026Updated this week
Alternatives and similar repositories for torchcomms
Users that are interested in torchcomms are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- NVIDIA NVSHMEM is a parallel programming interface for NVIDIA GPUs based on OpenSHMEM. NVSHMEM can significantly reduce multi-process com…☆563Updated this week
- NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer☆195Feb 11, 2026Updated 5 months ago
- PyTorch Single Controller☆1,060Updated this week
- A storage solution for PyTorch tensors with distributed tensor support.☆81Jul 17, 2026Updated last week
- A NCCL extension library, designed to efficiently offload GPU memory allocated by the NCCL communication library.☆111Dec 17, 2025Updated 7 months ago
- Simple, predictable pricing with DigitalOcean hosting • AdAlways know what you'll pay with monthly caps and flat pricing. Enterprise-grade infrastructure trusted by 600k+ customers.
- Fault tolerance for PyTorch (HSDP, LocalSGD, DiLoCo, Streaming DiLoCo)☆526Jul 16, 2026Updated last week
- MSCCL++: A GPU-driven communication stack for scalable AI applications☆542Updated this week
- Distributed Compiler based on Triton for Parallel Systems☆1,498Updated this week
- UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g…☆1,469Updated this week
- Perplexity GPU Kernels☆591Nov 7, 2025Updated 8 months ago
- An experimental communicating attention kernel based on DeepEP.☆34Jul 29, 2025Updated 11 months ago
- ☆378Jan 28, 2026Updated 5 months ago
- Accelerating MoE with IO and Tile-aware Optimizations☆732Jul 4, 2026Updated 3 weeks ago
- A Python-embedded DSL that makes it easy to write fast, scalable ML kernels with minimal boilerplate.☆913Updated this week
- Virtual machines for every use case on DigitalOcean • AdGet dependable uptime with 99.99% SLA, simple security tools, and predictable monthly pricing with DigitalOcean's virtual machines, called Droplets.
- Perplexity open source garden for inference technology☆604May 27, 2026Updated last month
- DeepXTrace is a lightweight tool for precisely diagnosing slow ranks in DeepEP-based environments.☆101Jan 16, 2026Updated 6 months ago
- ☆170Dec 27, 2024Updated last year
- A Quirky Assortment of CuTe Kernels☆1,066Updated this week
- Triton-based Symmetric Memory operators and examples☆106May 15, 2026Updated 2 months ago
- NVIDIA Inference Xfer Library (NIXL)☆1,147Updated this week
- A lightweight design for computation-communication overlap.☆242Jan 20, 2026Updated 6 months ago
- ☆60Feb 24, 2026Updated 5 months ago
- A fast communication-overlapping library for tensor/expert parallelism on GPUs.☆1,345Aug 28, 2025Updated 10 months ago
- AI Agents on DigitalOcean Gradient AI Platform • AdBuild production-ready AI agents using customizable tools or access multiple LLMs through a single endpoint. Create custom knowledge bases or connect external data.
- ☆45Oct 15, 2025Updated 9 months ago
- Tutorial Exercises and Code for GPU Communications Tutorial at HOT Interconnects 2025☆32Oct 22, 2025Updated 9 months ago
- Tile-based language built for AI computation across all scales☆174Jun 16, 2026Updated last month
- An experimental implementation of compiler-driven automatic sharding of models across a given device mesh.☆89Updated this week
- MSLK (Meta Superintelligence Labs Kernels) is a collection of PyTorch GPU operator libraries that are designed and optimized for GenAI tr…☆121Updated this week
- mKernel: fast multi-node, multi-GPU fused kernels☆254Jun 21, 2026Updated last month
- Venus Collective Communication Library, supported by SII and Infrawaves.☆152Jun 24, 2026Updated last month
- a size profiler for cuda binary☆71Jan 15, 2026Updated 6 months ago
- Low overhead tracing library and trace visualizer for pipelined CUDA kernels☆137Jul 17, 2026Updated last week
- Serverless GPU API endpoints on Runpod - Get Bonus Credits • AdSkip the infrastructure headaches. Auto-scaling, pay-as-you-go, no-ops approach lets you focus on innovating your application.
- Distributed MoE in a Single Kernel [NeurIPS '25]☆273May 5, 2026Updated 2 months ago
- AMD RAD's multi-GPU Triton-based framework for seamless multi-GPU programming☆193Updated this week
- Open ABI and FFI for Machine Learning Systems☆435Updated this week
- Checkpoint-engine is a simple middleware to update model weights in LLM inference engines☆977Jul 4, 2026Updated 2 weeks ago
- NCCL Profiling Kit☆155Jul 1, 2024Updated 2 years ago
- Allow torch tensor memory to be released and resumed later☆260Updated this week
- High-performance distributed data shuffling (all-to-all) library for MoE training and inference☆123Mar 7, 2026Updated 4 months ago