Humming is a high-performance, lightweight, and highly flexible JIT (Just-In-Time) compiled GEMM kernel library specifically designed for quantized inference.
☆301Oct 22, 2026Updated this week
Alternatives and similar repositories for humming
Users that are interested in humming are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- DeepStack: Facilitating Co-Design Exploration of 3D DRAM-Stacked Accelerators for Distributed LLM Inference. Includes the MICRO 2026 AE a…☆65Oct 5, 2026Updated 2 weeks ago
- ☆520Aug 26, 2026Updated last month
- Reference implementation and examples of the CuTe Layout representation and algebra.☆380Sep 30, 2026Updated 3 weeks ago
- ☆116Aug 5, 2026Updated 2 months ago
- ☆130Updated this 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.
- ☆258Oct 14, 2026Updated last week
- NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer☆207Feb 11, 2026Updated 8 months ago
- ☆65Apr 26, 2025Updated last year
- ☆18Jun 19, 2026Updated 4 months ago
- CuTeDSL tutorials, tools, autotuner, profiler, etc.☆47Jun 27, 2026Updated 3 months ago
- DeeperGEMM: crazy optimized version☆85May 5, 2025Updated last year
- High Performance LLM Inference Operator Library☆1,184Sep 7, 2026Updated last month
- Agent-friendly GPU profile-query CLI☆128Oct 4, 2026Updated 2 weeks ago
- DeepSelect: TopK kernels for DeepSeek Sparse Attention (DSA) and Samplers☆495Sep 30, 2026Updated 3 weeks ago
- Bare Metal GPUs on DigitalOcean Gradient AI • AdPurpose-built for serious AI teams training foundational models, running large-scale inference, and pushing the boundaries of what's possible.
- ☆304Sep 5, 2026Updated last month
- FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.☆1,166Sep 4, 2024Updated 2 years ago
- ☆38Aug 4, 2026Updated 2 months ago
- Perplexity GPU Kernels☆610Nov 7, 2025Updated 11 months ago
- Mixture-of-experts (MoE) training megakernel for NVL72s☆625Oct 1, 2026Updated 3 weeks ago
- ☆32Jul 2, 2025Updated last year
- Accelerating MoE with IO and Tile-aware Optimizations☆781Aug 29, 2026Updated last month
- Kernel Design Agents (KDA) is a agent-centric workflow to write high-performance CUDA Kernels.☆1,431Updated this week
- CUDA kernels for linear attention variants, written in CuTe DSL and CUTLASS C++.☆557Updated 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.
- Production-ready MoE load balancing via real-time expert replication☆258Oct 15, 2026Updated last week
- ☆68Feb 24, 2026Updated 7 months ago
- High-performance LLM operator library built on TileLang.☆196Updated this week
- FA4-based Relative Attention Kernel developed by TML and Colfax☆18Sep 11, 2026Updated last month
- Unified tracing profiler and visualizer, one timeline from CPU to GPU/HES to in-kernel zones☆149Sep 21, 2026Updated last month
- mKernel: fast multi-node, multi-GPU fused kernels☆296Updated this week
- A lightweight design for computation-communication overlap.☆250Jan 20, 2026Updated 9 months ago
- MoonEP: A Perfectly Balanced Expert Parallelism Library via Dynamic Redundant Experts☆1,189Oct 12, 2026Updated last week
- ☆98Feb 5, 2026Updated 8 months ago
- GPU virtual machines on DigitalOcean Gradient AI • AdGet to production fast with high-performance AMD and NVIDIA GPUs you can spin up in seconds. The definition of operational simplicity.
- QuTLASS: CUTLASS-Powered Quantized BLAS for Deep Learning☆204Sep 2, 2026Updated last month
- TensorIR is a lightweight NVIDIA-owned MLIR compiler frontend for expressing tensor computations and lowering them to NVIDIA CUDA Tile IR…☆138Oct 15, 2026Updated last week
- Boosting 4-bit inference kernels with 2:4 Sparsity☆101Sep 4, 2024Updated 2 years ago
- NaiveRT: a self-contained TP8 inference runtime for the Naive N series on eight Blackwell GPUs with UMMA support. One persistent CUDA gra…☆30Sep 27, 2026Updated 3 weeks ago
- An agent harness that compiles a model into one provably-correct, self-retargeting CUDA megakernel and self-tunes it past cuBLAS at batch…☆154Sep 18, 2026Updated last month
- Compact and Agent-Native MoE Training System☆357Updated this week
- Tilus is a tile-level kernel programming language with explicit control over shared memory and registers.☆495Updated this week